[ NNSquad ] Re: Tim Berners Lee, inventor of WWW, speaks out in favor of Net Neutrality

[Barry Gold <BarryDGold@ca.rr.com>]

On 9/16/2010 8:54 AM, Lauren Weinstein wrote:
> Tim Berners Lee, inventor of WWW, speaks out in favor of Net Neutrality
>
> http://bit.ly/dp1ovm  (Guardian)

I note that Berners-Lee says:

"...they think that because you want access to data you must be an 
executive! And there's no in between."

 and

"The fact is that even small amounts of data are very effective for 
connecting people."

 and

"I'd like people enrolled in a low data package by default."

I think Berners-Lee, at least, recognizes that not all plans need to be 
"all you can eat," nor does everybody need a high-speed connection.

That corresponds pretty closely with my view: charge different amounts 
for different speeds, allow for plans that limit the amount of data you 
can transmit per month -- but within those limits, all data transmitted 
via IP is equal(*).

I can't buy into the "there is no scarcity" argument.  Even if there 
_is_ a bunch of dark fiber out there, it costs money for power to run 
the amplifiers and routers, money to repair the cable when it stops 
working, money to manage the networks so they don't fall apart under the 
inevitable DDOS and other attacks.(+)

So you need for providers to make a profit, so they will have an 
incentive to build more cable, construct higher-speed "backbones", etc. 
 At the same time, you need to keep their profit motive from making the 
net unusable for others.

There are basically two things I am worried about:

1. Cross-subsidization, where the telco or cableco jacks up the price of 
high-speed internet -- even if they have to "pay" themselves the higher 
price, they can absorb that "loss" (remember, it's funny-money) until 
they have driven their competitors -- Hulu, YouTube, Vonage, etc. out of 
business.

2. The "supermarket" business model.  Supermarkets do a wonderful job of 
getting popular items to us at a reasonable price, and also carry an 
amazing variety of different items.  But if you want your product at or 
near eye-level, you have to _pay_ the market to put it there.  And in 
some cases, you may have to pay the supermarket just to put your product 
on the shelves at all.  Consider "gift" debit cards.  You can go to the 
store and buy a card, pre-loaded with $100.  A really neat way to give 
somebody a gift, or just to convert money in your pocket into a form 
that you can use on the web.(@)  But that $100 card will cost you $106 
at checkout.  What?  You're giving them the use of _your_ money, for 
free, until you get around to spending it.  And if you don't spend it 
within 6 months, they start charging you $5/month, so pretty soon it's 
all gone.

You gotta figure that there's not much cost to running those cards -- no 
statements to mail, almost all the work is done by computers.  So 
logically, there should be other companies offering these cards for less 
money, and the price should get pretty low.

But it doesn't happen that way; my guess is, the reason is that the card 
issuers have to pay the markets a premium just to get their card on the 
rack.  So there's not much competition, and the few Visa/
Mastercard/Amex gift cards available all carry these high fees.

That's what I'm afraid could happen to the Internet: if you want your 
wonderful new idea to be available to the average person, you have to 
not only spend money to advertise it, but you have to pay _every_ ISP 
(or at least, all the ones with more than a couple of hundred 
subscribers) just to have it be _available_ at a reasonable speed.

I _think_ we can get by with reasonable regulation (no preference for 
one packet over another, and _some_ guard against cross-subsidization), 
but it's possible that we will have to go to the utility model, with 
regulated rates etc. -- or the _other_ utility model -- you can charge 
anything you want, but you have a _single_ business: carrying packets. 
No other product lines on the same cable to cross-subsidize.

(*) If ICMP ever proves useful for something other than tracing the 
route to another host, I could see allowing a higher priority for _a 
limited amount_ of ICMP packets: you can have 1 meg of ICMP / month, 
after that they get routed like normal packets.

(+) It's a little bit like, "there is no excuse for starvation, we have 
plenty of food to feed everybody in the world."  That's true.  But if 
you feed people without them having to work for it, there will soon be 
more people, and in at most a few decades there will be a food shortage 
again.  The way to solve starvation is to get rid of the obstacles that 
keep people from feeding themselves -- government edicts that force 
nomadic peoples to stay in one place, where their goats promptly eat up 
all the grass and create a mini-desert -- fighting and other problems 
that keep food from reaching those who need it -- etc.  Then the 
increasingly low (inflation adjusted) price of food will solve the 
problem all by itself.  (Mind you, I contribute to groups like Heifer 
International, as a way of getting people through the current crises, 
but I also recognize that we need a real long-term fix.)

(@) If ever there was a good case for patenting a business model, this 
should be one.  It's a simple, yet brilliant concept, and could probably 
have been implemented 25 years ago.  But it wasn't, until somebody 
thought of it.