- To: slug@xxxxxxxxxxx
- Subject: Re: ADSL is not a direct connection to *The*Internet* (Was: ADSLrant (was: Re: [SLUG] Telstra Cable Anouncement))
- From: Roland Turner <raz@xxxxxxxxxxxxxx>
- Date: Tue Aug 8 18:26:18 2000
- Organization: -
Andrew Macks wrote:
> On Tue, 8 Aug 2000, Roland Turner wrote:
> > The local cable loop _can_ be a bottleneck, but the pressures that bring
> > this about are commercial, not technical. Likely circumstances include:
> > - Optus has grossly underprovisioned and is being swamped with business.
> > Oops. Same can happen with ADSL. This is a commercial error.
> They are handling the customers, and their network is not grossly
> bottlenecked due ot the demand. So I don't believe this to be a problem.
Sorry, I was presenting hypotheticals. I could perhaps have worded it
> > - Optus is failing to adequately penalise users who are abusing the
> > service. Oops. Same can happen with ADSL. This is a commercial error.
> Having known four people personally who have lost Optus@Home accounts due
> to over-usage and having seen the e-mails come through this list of server
> scanning and warning, and deleting of accounts on this basis. They are
> dealing with it.
> > - An area is serviced by Optus cable, but not by either of the Bigpond
> > broadband offerings. Optus exploits its monopoly. Same can happen with
> > ADSL. This is commercial behaviour.
> This isn't a monopoly. Optus haven't done anything to stop Bigpond coming
> into the area. If anything, Bigpond have a monopoly because they won't
> allow Optus to use their underground cable housing infrastructure, and if
> it weren't for the ACCC allowing Optus to use it in areas where overhead
> cables are forbidden, imagine that monopoly! It's not Optus' practises
> that have caused the monopoly, it is by chance.
Again, a hypothetical.
To clarify language though: the recent focus on a certain high-profile
anti-trust case in the USA has perhaps blurred perceptions. A monopoly
is simply a sitatuion where customers are, for whatever reason, unable
to choose to do business with a competitor. In areas serviced by Optus
cable that are not serviced by either Bigpond broadband offering, Optus
can be said to hold a monopoly. (*) An entirely seperate concern is the
use of anti-competitive practices, in any situation, to pursue business
advantage. Several such practices are rendered illegal by the Trade
Practices act here. Additionally, in the USA, section 2 of the Sherman
anti-trust act renders the use of anti-competitive practices to create
or sustain a monopoly illegal. It is not the case that the existence of
a monopoly implies the use of anti-competitive practices to create it.
(* A big problem with identifying a monopoly is identifying/defining the
market. If you broaden the market definition to include the demand for
high-capacity digital telephony services and/or satellite, then you
would suddenly observe that Optus did not have this hypothetical
monopoly. I suspect that, for most customers, these options are not
attractive, consequently in (hypothetical - I don't know whether they
even exist) areas that are served by Optus cable but not by either
Bigpond offering, Optus holds a monopoly. This does not imply that Optus
used anti-competitive practices to establish this monopoly, or that it
has done any other unreasonable thing, merely that customers in those
hypothetical areas can't choose to use any other provider of an
> I don't think people here are mindless enough to assume that. The point
> that was trying to be made was that Cable is a shared access scheme whilst
> ADSL is not. Of course there are other dependencies further up the chain,
> but that's a different concern.
I have the impression that Dazza's point is subtly different: his
viewpoint appears (to me) to be that if you choose cable you will get
poorer service because you are using a shared access medium whereas if
you choose ADSL you will get better service because you aren't. (This is
quite different from arguing that cable is a shared medium and ADSL is
not. This is a question of fact and, clearly, we are in agreement about
it.) As customers of any of the three services in question do not get to
mix and match, then this concern cannot be meaningfully seperated from
concerns about the shared media immediately upstream. My argument is
that the service that you'll experience will be the result of commercial
pressures, not technical limitations. Dazza's argument is that the
service that you'll receive will be the result of technical limitations,
not commercial pressures.
> > Major ISPs have no excuse for mail service outages at all. For the sake
> > of putting a number on it, anything less than five nines uptime is
> > inexcusable. (99.999% uptime implies no more than about five minutes of
> > downtime per year.)
> Optus@Home and Bigpond don't seem to know how to handle a large mailbase.
> I'm not saying it's easy, I'm just saying it could be handled better than
> dumping it all into an unstable network.
I think that we are in agreement here? (I'm not saying that it's easy. I
am saying that it could be handled a lot better.)
> > > HFC is an inferior solution to ADSL - simply because of the shared
> > > bandwidth. If I can get ADSL to my place, I'm going for it.
> > These are two seperate assertions. The latter may be true, and you are
> > free to spend your money as you see fit. The former is false, simply
> > because of the shared bandwidth. The differences are commercial, not
> > technical.
> Differences are technical. HFC upstream is very very much shared. The
> downstream is a little more segmented. ADSL is shared amongst you up and
> downstreams, but not amongst other users at the delivery level of the
The actual sharing arrangements are of no interest to users of the
service who are concerned with better performance. For these users, as
long as the service is performing well, they are happy.
Let me try this a different way: Suppose that you own a cable network
and that you sell Internet access through it. Suppose also that there is
a competing ADSL provider serving the same area. Why would you even
_permit_ your cable to become saturated? Would you not go out of your
way to ensure that users who were abusing the service were stopped and
that the segments were kept small enough to keep providing the capacity
that the customers are paying for so that they did not run to your
competitor? This is no different to provisioning any other kind of
growing network, you grow it as needs dictate. Your ADSL competitor
wouldn't install a 2Gbps feed to a DSLAM that is only expected to feed
50 customers in the next 12-18 months, similarly you wouldn't split
cable segments until demand required it. The fact that these unsplit
cable segments exist doesn't mean that they will exist eternally.
Splitting the cable segments as neccessary over time is a part of the
business model, as is enlarging upstream capacity. In order to claim
that the cable provider will not do this, it is neccessary to assert
that the business is not willing to grow its expenditure as sales
increase (_very_ few business have this characteristic, no competitive
telcos do) or that it is difficult or infeasible to grow the
infrastructure (it is not). (This last assertion is the same as saying
that the cable Internet providers have poured huge amounts of cash into
a business model that does not allow them to sell more as customers
start wanting to buy more. Again, very few businesses have this
characteristic, no competitive telcos do.)
So, to the assertion that cable uses shared media to your door where
ADSL does not, I of course agree, it is a matter of fact.
To the assertion that ADSL is therefore a better choice than cable for
people wishing to have a broadband connection, I respond, again, that
there is no causal relationship here (between technology choice and
performance, from the standpoint of Joe Average Punter). In a given
geographic area at a given point in time, any one of the three is
capable of being the best performing. This leadership can even change
over time. The determinant of what service you will receive from any one
of them is commercial, not technical. No facts to the contrary have been