Re: Should we oppose the Data Superhighway/NII?

Mike Godwin <mnemonic@eff.org> said:
>Perry writes:
>> Seems to me that bandwidth is
>> going to be nearly free in both directions in a few years whether
>> government intervenes or not.
>
>I agree about the potential for it to be free, but, I gotta tell you, the
>monopolists running the cable systems in this country have no inclination
>to share that nearly free bandwidth with you, even if you're willing to
>pay for access to it.

I hate to disagree, considering that I prefer to agree with the philosophy
here, but it *can't* work that way, regardless of what we wish.

The problem is that bandwidth is a highly limited resource, just like
real estate is a limited resource. Eventually we will complete saturate
network bandwidth no matter what technology is used. This has been discussed
in various forums for many years. Once optical fiber optic bandwidth
peaks, you have to move to ultraviolet for greater channel capacity.
Then that is exhausted, and we will continue pushing...gamma ray bandwidth
fiber optic (or line of sight transmission) will eventually be a target,
despite its extreme difficulties even in theory.

At the same time we will be laying fiber and raising dishes to beat the
band. But no matter how well all that goes, we will *very* quickly reach
a saturation point of facilities as each new technology is introduced.

These days it's easy to be optimistic, because bandwidth is growing
geometrically. The problem is that there is no way in hell that that
trend can continue indefinitely. One or two decades hence we will saturate
theoretical limits.

Bandwidth is and will always remain a scarce and precious resource.

On the other hand, if you mean "slow channels by comparison with state of
the art channels," then yeah, *that* may as well be free at any given
point. Right this instant one could make an argument for 110 baud
channels being free.

>In order to get to a world in which free markets can meet our demand for
>high-bandwidth connectivity, we have to dig ourselves out from the 
>market-failure position we're in now. And because government is part of
>the problem, changing government policy is part of the solution. So,
>that's one of the major thrusts of EFF's NII policy. 

I agree, but this seems to be a subject change.
	Doug

Doug Merritt writes:

> Mike Godwin <mnemonic@eff.org> said:
> >Perry writes:
> >> Seems to me that bandwidth is
> >> going to be nearly free in both directions in a few years whether
> >> government intervenes or not.
> >
> >I agree about the potential for it to be free, but, I gotta tell you, the
> >monopolists running the cable systems in this country have no inclination
> >to share that nearly free bandwidth with you, even if you're willing to
> >pay for access to it.
> 
> I hate to disagree, considering that I prefer to agree with the philosophy
> here, but it *can't* work that way, regardless of what we wish.
> 
> The problem is that bandwidth is a highly limited resource, just like
> real estate is a limited resource.

Doug, I think you may be under the impression that we're talking about a
single fiber-optic or coax cable. You can have a single (or double or
triple) infrastructural network, but add bandwidth to each one. I think
the notion of "scarcity" doesn't apply to cable any more than it applies
to personal computers.

> Eventually we will complete saturate
> network bandwidth no matter what technology is used. This has been discussed
> in various forums for many years. Once optical fiber optic bandwidth
> peaks, you have to move to ultraviolet for greater channel capacity.

Or you add a new cable. Not hard. Nothing I have ever read has suggested
that "scarcity," as that term is normally used in reference to a resource,
applies in any meaningful way to cable. I believe that Perry's prediction
is closer to the truth than yours.


--Mike

Doug Merritt says:
> I hate to disagree, considering that I prefer to agree with the philosophy
> here, but it *can't* work that way, regardless of what we wish.
> 
> The problem is that bandwidth is a highly limited resource, just like
> real estate is a limited resource. Eventually we will complete saturate
> network bandwidth no matter what technology is used.

Lets see whether this is reasonable.

A single fiber optic strand has enough capacity in theory to carry the
equivalent of every call made in the U.S. during the peak capacity
utilization period on Mother's Day. A single fiber can carry more data
than can be transmitted by the entire radio spectrum from low
frequency AM to Ku band satelite. Thats bandwidth for literally
thousands of simultaneous video signals.

Using switching technology rather than shared access LAN style
technology, every person in the world could concievably be sending and
receiving that much at once.

I don't know about you, but I personally can't produce more than 750
simultaneous videos at once for network distribution, so I suppose I'm
uninteresting, but even the people who can do more than that are
likely going to be fine. If they aren't, well, I suppose they could
get TWO fibers coming into their home, or maybe even TEN or ONE
HUNDRED if necessary.

> These days it's easy to be optimistic, because bandwidth is growing
> geometrically. The problem is that there is no way in hell that that
> trend can continue indefinitely. One or two decades hence we will saturate
> theoretical limits.

I suspect that we have a wee bit longer to go than that. When people
start faxing themselves regularly we may have to go to slightly more
exotic technologies.

Perry

In article <9311111430.AA28017@snark.lehman.com> pmetzger@lehman.com writes:
>A single fiber optic strand has enough capacity in theory to carry the
>equivalent of every call made in the U.S. during the peak capacity
>utilization period on Mother's Day. A single fiber can carry more data
>than can be transmitted by the entire radio spectrum from low
>frequency AM to Ku band satelite. Thats bandwidth for literally
>thousands of simultaneous video signals.

"All the world's comms needs can be met with a single fibre"

contrast this with a famous quotation from history...

"All the country's computing needs can be met with a single computer"

:-)

G
(It'll be a nice one to quote to my grandchildren in years to come...)

Graham Toal says:
> In article <9311111430.AA28017@snark.lehman.com> pmetzger@lehman.com writes:
> >A single fiber optic strand has enough capacity in theory to carry the
> >equivalent of every call made in the U.S. during the peak capacity
> >utilization period on Mother's Day. A single fiber can carry more data
> >than can be transmitted by the entire radio spectrum from low
> >frequency AM to Ku band satelite. Thats bandwidth for literally
> >thousands of simultaneous video signals.
> 
> "All the world's comms needs can be met with a single fibre"
> 
> contrast this with a famous quotation from history...
> 
> "All the country's computing needs can be met with a single computer"

You didn't understand the point. Not a SINGLE fiber. Each person would
have their own fiber into a switched fabric the way everyone has their
own phone line into a switched fabric. Each person would have all that
capacity TO HIMSELF. If he needed more, he could get two, or even ten,
not that he'll need more than one.

Perry