Re: [ZDNet] Broadband disappoints customers.

Mike Masnick (mike@techdirt.com)
Tue, 08 Dec 1998 23:08:59 -0800


I know I've had this conversation before with Ian, but I'm still not
totally clear on this. I also admit it's been a while since I looked at
this market (I actually did a paper on xDSL v. Cable Modems over two years
ago).

At 02:26 PM 12/8/98 -0800, Ian Andrew Bell wrote:
>Architecturally, CableModems are a much better technology than DSL for
>several reasons:
>
>1) CableModems are more like the internet in that self-organizing
> engineering principles drive closer to the home. This means that
> some of the economies generally associated with internetworking
> propagate to the residential gateway as well. DSL is still based
> on the tried-and-true (but cost prohibitive) circuit-switched model
> of service delivery, versus the shared pipe (and ultimately the
> optimization) of CM's.

But doesn't this also mean that DSL gives you a more "guaranteed" level of
service than CM? The shared bandwidth is fine when the usage is low, but
when everyone's got their CMs and using them for high bandwidth
applications then it causes trouble, right? I realize this is only a piece
of the puzzle (and distance traveled, but it's still worth noting). Is
this correct?

>2) CableModems allow the service provider who already delivers CCTV
> services to enter two brand new markets -- data services, and
> telephony. DSL is simply an adaptation of already existing data
> connectivity but is still not adequate, in terms of bandwidth, for
> doing CCTV. DSL actually cannibalizes hefty Frame Relay, ISDN,
> and T1 data service revenues from TelCos.

I agree with this, but I tend to believe the telcos aren't that stupid.
They realize that they need to cannibalize their own services or the cable
companies will...

>3) Cablemodems are less sensitive to the loop distance vs. bandwidth
> degradation and variability that DSL suffers from.

No argument here. However, I'll admit that this is the first time I've
seen this argument. DSL is obviously based on the loop size, so of course
there's degredation as the loop gets bigger... The counter to this is
simply the shared vs. guaranteed argument above.

Other things I remember hearing about:
- The cost to upgrade cable systems is a hell of a lot more than they can
afford (don't know how true this is, and smacks of telco propaganda).
- The percentages of people who have copper to the homes vs. coax. I
belive only about 60% currently have cable.
- I vaguely remember some stat about which company you trust more, and it
went on the side of telco's, but most people (I imagine) hate both.
- Internal location of cable and copper. Most computers aren't near
cable. This is not a very strong argument, obviously...

I definitely see where you're coming from, but almost all of the things
I've read have felt differently over the long run. Most folks I know who
have either, love it. My guess is that both will be accepted in certain
areas. At some point, they'll have to be comparibly priced, and it'll be
difficult knock one out of the market entirely.

Those are just my (very tired) thoughts, based on fairly old data and
research. Any info you've got that would help explain your side would be
great...

-Mike