WORTH READING "not rely on carriers" ????

[David Farber <dave () farber net>]

Begin forwarded message:

From: mo () ccr org (Mike O'Dell)
Date: October 29, 2008 8:03:06 PM EDT
To: dave () farber net
Subject: "not rely on carriers" ????


this is an interesting tale,
but if you don't "rely" on carriers,
what exactly do you plan to use for wires
and fiber?  surely you are not suggesting
simply re-implementing it all somehow
(or even just the parts you like).

people seem to forget what a carrier really is.

first and foremost, it's a business entity which can successfully
raise enough money to build and operate communication
infrastructure. that involves cutting deals for rights of way,
negotiating with regulatory bodies at the local, state, national,
and international levels, deploying a lot of hardware, and
then owning and operating that hardware.

note that nothing I said up to this point says anything
about "business models". whether a carrier's service
is selling IRUs for dark fiber, access to lambdas,
point-to-point 10gigEthernet, or any other interface
exposed by transport equipment is interesting only
to the degree it determines whether they sell what
you want to buy.

the most important fact is:

carriers move bits.

that's it. that's the killer app
for true carriers.

there are ISPs which provide global delivery of IP packets
by using the bit-moving services of carriers.
they add other infrastructure on top of the bit-pipes
supplied by carriers and generate Internet Service from it.

Note: it is not uncommon to find a carrier and an ISP
squished into the same higher-level corporate entity
(usually indicated by pervasive logo stains), but there
*are* two different entities inside there. it's the
upper-level logo who is blenderizing the "business models".

these upper-level logos often peddle other wares, too,
but they are of concern only to the degree they get in
the way of using the carrier or ISP services.

this picture is obviously simplified - some carriers specialize
in escorting bits across ocean passages. the dollars required
to do that magic are considered large by "dry-land" carriers,
and dry-land carriers can spend multiple hundreds of millions
of dollars on deployment with surprising ease. why?
the world is a big place, and as we used to say,
"The fiber am where the fiber is." it doesn't go
everywhere and cannot go everywhere, so some places
are more equal than others when it comes to delivering
different kinds of bit-pipe service. the same is true
for every other bit-pipe technology.

so what i really don't understand is what part of this
we can do without?  certainly the brain-damaged blenderizing
of business models could be jettisoned, but i don't see
how you build anything like what we now know as
"The Global Big-I Internet" without fiber and submarine
cables and metro distribution and local tails.
and like it or not, unless you intend to completely
re-implement all of that, you are going to need the
facilities of "carriers".  true, it would be nice if
they would find a way to be happy with their true
killer app, but that doesn't make those bit pipes
less necessary.

        -mo





-------------------------------------------
Archives: https://www.listbox.com/member/archive/247/=now
RSS Feed: https://www.listbox.com/member/archive/rss/247/
Powered by Listbox: http://www.listbox.com