PSH, I've recently refrained from commenting on ATHM's bandwidth
attributes on other threads, but I guess you've put it squarely to me, so it's
time to address the matter.
"What are your thoughts on @HOME being a pure distributor of
bandwidth with no direct role in providing content? "
The short answer:
ATHM doesn't have any bandwidth to sell. At least not where it counts,
in the Last Mile. And, where they do have it to sell, on their backbone,
it's available from others with little of no constraints. And lastly,
their servers and caching are to the best of my knowledge not open
standards based to the extent that they would present a friendly
interface to the WWW. Instead,they are proprietary to meet the needs
of tenants on ATHM's intranet. Comments and corrections on my take
of these last several issues would be welcome.
In the LM, ATHM is a "tenant," riding over the MSOs' facilities. And,
unless they are going to sub-lease their own limited bandwidth, I am
hard pressed to come up with another supply that they might have to
sell.
The longer answer:
I believe it was you (it could have been someone else) on the ATHM
thread who suggested that ATHM should renounce their contractual
obligations to the MSOs - the same MSOs who own them, for the most
part - and sell their bandwidth to other SPs.
Assuming such an extrication could take place legally or assuming that
such could be negotiated, the argument suggests that ATHM would be
free to open up their abundant supply of bandwidth and sell access to
other ISPs to attach to their last mile subscribers, much the same way as
you are suggesting here.
In the process, the argument continues, ATHM would reap the revenues
from many more subscribers than they are able to bring online now, or
more than they are experiencing at the present time under their own
branding. Implied in this observation is the fact that ATHM's rollout is
taking far too long, and that the competition continues to outpace ATHM
while increasing their share of the market.
Some fundamental questions:
Where would ATHM get this bandwidth to sell, once they have
scorned their present owners? (assuming that is the case. Of course, in the
unlikely event that the owners buy into this idea, then that's something
else);
[ii] Who would perform all of those additional installations, if ATHM did
have the bandwidth to sell, and if they could attract the additional tiers of
ISPs beneath them who would settle on a second (third?) fiddle role?;
and,
[iii] Why would these installations take any less time to complete than
ATHM's own offerings?
The last mile does not belong to ATHM, in any event. Instead, the LM
plant is owned by the MSOs and others.
The only bandwidth that ATHM has to its name which it could make
available to the other ISPs [who would become customers of ATHM in
the process] would be the dual OC-48 backbone rings, which,
ostensibly, lie on the Interior of ATHM's cloud, behind the many clusters
of MSO head ends. These rings tie into the ATHM server and cache
farms and connect to each head end cluster.
Are these the sources of bandwidth which you are referring to? The
interior rings? Because, if they are not, I am hard pressed to come up with
any other source of bandwidth that is, at least by popular notion, solely
theirs.
[[And here, I have a question which has never been fully answered to my
satisfaction. Do these OC-48 rings which T is putting in place for ATHM
truly belong (owned outright) solely by ATHM [the corporation ATHM],
or do these constitute a cloud which is owned by the individual MSOs, on
a shared basis, with each MSO owning a percentage share in them, and
each MSO possessing sovereign rights for their own traffic, as well? If
someone could answer this with some authority, I'd appreciate it.]]
Another point to consider is that the MSOs also own the head ends and
the optical elements which deliver & receive optical (increasingly,
DWDM) flows from the hybrid fiber/coax (HF/C) distribution and feeder
networks out to the neighborhoods.
Head ends are not assets which lend themselves very easily to
disassembly and allocation between normal program TV services (or even
cable telephony services, in some instances) and those related to Internet
service provider interests, since there are many system elements in head
end assemblies which are common to both or all three service segments
(Program TV, Internet, and Voice).
Thus far, I've only addressed the physical layer components. It gets more
entangled when you take into account the delivery of actual "services,"
and the inclusion of a common billing and management platform to
address all of the above, along with the electronic bonding requirements
which exist to other xLECs.
The latter operations support systems (OSSes) play into synchronizing
inter-carrier directory functions, voice network signaling, billing,
network management, etc. And the foregoing should also serve to partially
answer the second part of your question concerning:
"Are there any other potential revenues in this model aside from being
the ultimate cable service provider... (eg voice-if the technical factors are
worked out?)"
Sure. Most of the inhibiting factors can be worked out from both purely
theoretical and technological standpoints. But with egos, politics and other
self-serving interests being what they are, none of them are likely to take
hold in a far reaching way during the foreseeable future, with the possible
exception of T's deployment of switched voice over cable, IMO.
In order for services to flourish, the MSOs must adopt one of two options
(perhaps a little of both, if they can pull that one off and still remain
whole). Those options would be
- the appearance of true orchestration (such as that offered by the PSTN
model) or
- the appearance of true chaos (which is offered by the Internet model)
In the first of these, orchestration would require a single mind set, a single
conductor waving a baton. This is principally the idea which was behind
CableLabs and its constituent MSO members, prior to the Internet's
influences expanding to the degree to which it has.
And I would venture to say that this is still behind the thinking at
CableLabs, although I think that they have a lot more on their minds these
days with the encroachment of the chaotic model outpacing users'
willingness to remain a part of the status quo for very long, than when their
earlier planning and blueprinting was more colloquialized and serene.
Well, I digressed again, as usual, but I think that I've addressed some of
the issues which would need to be taken into account, if your idea were
ever taken to the implementation level.
Regards, Frank Coluccio
|
