Teddy, I can see where some of my message caused confusion. I wrote in
a way that assumed too much of readers, and I should have been clearer.
Teddy: "... there are a two statements is your post that left me
significantly perplexed. They seem almost contradictory:.. I don't recall
you using the term “convergence freaks” before.
The IETF and its decades-old methods of innovation through
experimentation and persistence are on one side of the equation. They
now face a new kind of challenge at this time from the ITU, I believe, and
the established carriers and their vendors who only recently came to
finally recognize the power, or should I say, the inevitability, of IP.
By "overshooting" in the direction of IP, I meant that FRO must maintain a
high level of awareness of where the other, and more established,
global players are taking this technology.
Are the larger players, too, going in the direction of a pure form of IETF
IP at one end of the spectrum, versus, are they taking it to a newer
incarnation of IP with inherited PSTN switching attributes, which is
what I see the incumbents fostering, at this time.
It's IETF innovation and consensus vs. PSTN emulation and extension,
the way I see it right now.
My point was that they should be careful to refrain from over-extending
in either direction, prematurely. As a new entity, FRO is particularly
vulnerable because they don't have existing platforms in either
regimen. What they do now could potentially result in either
(a) being stranded in a world of underutilized IETF RFCs, while the
incumbents have fashioned something altogether different at their gateway
and accounting toll booths, or, on the other hand,
(b) if they remain overly conservative in the use of switched platforms,
even at their CLEC level, they could find themselves in a similar dilemma
from the other direction if IP should be adopted more rapidly by the
larger players.
Beyond the question being "either / or," it's also a matter of what kind and
to what degree, since all carriers will face the need to meet both new tech
and old tech classes of users (see Isenberg's article referenced later on).
"Convergence freaks" was perhaps my worst selection of terminology in
the preceding message, because it very likely would lead one to conclude
that I was referring to the traditional Internet crowd (as I surmised you
had interpreted it). I was not.
Instead, I was referring to the hundreds (thousands?) of newly formed
companies and inter-working group consortia/working groups, who are
made up of non-IETF affiliated entities. Okay, some may actually have
token representation through sign up memberships, and in some cases,
granted, some of the larger established vendors have actually been at the
forefront of IETF activities, but with a more traditionally IETF slant in the
past than they do at this time.
Many of these organizations aspire to bring their own flavors of integrated
*whatever* types of services to end users through *whichever* means
they can, irrespective of the consequences to global reach. In many ways,
in fact, vendor differentiation in these areas amounts to the antithesis of a
primary principle upon which the Internet has thrived: Consensus.
Some of these areas are out of control right now, and will wind up going
nowhere but south. In this sense, I could be referring to many of the
startups in the VoIP space, as an example.
The only way that many of these newco's have to get around their own
limitations is through the use of almost- hermetically sealed islands of
connectivity, because they are not aligned with globally accepted (read:
IETF sanctioned) Internet Telephony RFCs and practices. Indeed, a
complete set of working RFCs in this space don't even exist, yet.
And the only means possible for them to get from one of these islands to
the next is through a discreet gateway function, aided by gate-keeping and
other remote server applications, none of which work very well, and more
of which do not meet with any kind of internationally recognized precepts,
except for the basic transport protocols endemic to IP.
The problem with this approach, aside from the obvious, is that there are
almost as many of these non-interoperable kludge types as there are
vendors, themselves.
Each of these initiatives, by small- and large- player alike, for the most
part seeks simply to emulate what the older PSTN has done for decades,
without bringing anything new to the table in the way of innovative features
or increased functionality. Worse, many of them are coming into the
market place with dreadfully worse quality than the legacy services which
they would replace, the ones which preceded them. The tradeoff on
quality or lack thereof seemed do-able at one point, as long as the delta in
per minute charges could justify the pain. But...
...what was once perceived by many of these upstarts to be low hanging
fruit is quickly becoming something much much less, instead. Where many
of them were loathed to discuss flat monthly rates just a year ago, we are
now seeing them change their views in this regard, and coming out with all
you can eat plans as a means of redeeming themselves. And at the same
time, the traditional carriers continue to neutralize almost all of the pricing
advantages that the earlier aspirants sought to arb, through ever-lower
per-minute pricing plans of their own.
At the same time, the incumbents have come to realize that they cannot
fight this trend, and they have begun to take on some of the newer
approaches themselves. Here again, however, when the incumbents get
involved, they tend to simply emulate what the PSTN has been doing all
along, only doing it more efficiently in selected segments of their
architectures, through the use of IP technology. And when I say
'incumbents' in this context I am not only referring to the carriers, but to
their vendors/suppliers, the regulators, and everyone else in the chain,
including large enterprise shops who tend to emulate the larger carriers, as
well.
To this end, David Isenberg paints a clearer picture (hey, he gets paid by
the word and has a staff to proof his work), as demonstrated in this July 1
excerpt from one of his articles in America's Network Magazine: (See my
earlier post in the Gilder Thread for another article and further url
information):
Message 10766034
Isenberg writes:
"Lucent's disruptive customers were creating markets that seem
irrelevant to today's telcos but, with Lucent's help, would expand from
below to engulf telephony-classic markets - this is my 'duck' test for
disrupters.... Surprisingly, Martin (LU Bell Labs CTO) focused on
similarities. He said, "Both the incumbents and the disrupters see the end
game of the market as very much the same. [They both want] a very
high-speed packet backbone, optics at the center, a variety broadband
access mechanisms [and Internet Protocol (IP) as] the predominant
application protocol." Describing a two-year-old difference between
telcos old and new, he declared, "Cost-effective circuit to packet - we
built it. That debate is over."
You observed, and asked:
"Surely you have not forgotten the plight of the hundreds of millions of
Internet users that now held captive by the PSTN (Public Switched
Telephone Network). I find it hard to believe that any LEC still believes
they can solve the problem by adding more switch trunk ports or
IMT's."...Where your statements above meant to imply that: 1. QoS
issues will push true convergence so far into the future that companies
now making strides in that direction may be forced to take a step or two
back. 2. The Incumbents still wield sufficient power to prevent (for
whatever reason) integration for the foreseeable future.
No, I haven't forgotten the plight. I'm encouraged that the
potential exists for their plights to be overcome through innovation
of the type that has been possible due to the processes that have been at
work by Internet proponents, if such work is permitted to flourish.
It is this last point, however, that I have some reservations about, as the
more powerful influences of the larger vendors and established carriers
are slowly, albeit, very subtly, supplanting the efforts of the IETF, I
believe.
I don't like to harp on the cost of entry into ITU-like forums, but it is a
salient and rather representative point that has to be made. The costs are
simply too overwhelming for smaller players to cope with, and presents a
new kind of barrier to representation for folks who would otherwise
participate at the IETF level.
Some of the newer forums which have been fashioned in this way have
dues structures which exceed $60,000 per annum, just for the privilege of
sitting at the table. Take the latest initiative which happens to follow this
model, for example, which I posted on the VoIP board last night.
"PUSH IS ON FOR A NEW IP BILLING STANDARD"
Message 10763396
The cost of participation in this latest working group [and of course you
cannot only be a part of this initiative, if you are not a part of at least a half
dozen or more others] is $25,000. This may not seem like much to LU
or CSCO, but let's not forget where the ideas that have brought the
Internet this far have come from in the past. Individuals, academics, small
IP shops, as well as from vendors who heretofore were sensitized to the
emerging openness of the 'net, but who I feel are at this time ready to
stem the tide of open innovation because the next round of big players will be
the incumbent GSTN carriers, and not necessarily the classical ISP
model, alone. The vendors, in effect, have shifted their focus to where
their markets are shifting, and in the process I fear that one of the enabling
characteristics of the 'net (its previous innovation model) will suffer,
accordingly.
It is the intent of this group to introduce into the Internet space the
accounting functions and capabilities of the older PSTN which IP stands
perched to replace. Think about that one for a moment.
While many of today's practitioners from all walks of the industry will
openly and readily agree that most of the costs associated with providing
voice services today can be traced to the accounting functions behind the
metering of usage and their attendant billing models, the PSTN
nevertheless stands ready to replicate these processes in their emerging IP
model, despite the IP model's obvious advantages which derive from the
absence of any such overhead encumbrances in its native form. So, what
is the alternative to billing? That's another thread.
My ravioli and meatballs are ready. Later.
Regards, Frank Coluccio |
