« mp3.com, we hardly knew you (II) | Main | New Year's in tokyo? »
I've got a piece in Wired about the push to shift ownership of networks back to customers. It build on the work of Alan McAdams and the IEEE to accelerate broadband deployment in the US. For a great collection of papers related to this, see this . The Burlington, VT, story is told here. And the Utopia project in Salt Lake City is another great example of this.
| Permalink | technorati
TrackBack URL for this entry:
http://lessig.org/mt/mt-tb.cgi/1075
Listed below are links to weblogs that reference Fiber to the People:
» Computers to 2011 from Planet P
Below are my predictions for personal computing (hardware and software) for the next 8 years. Predicting even where computing will be is getting more difficult as there more and more disruptive advancements just wating to be unleashed - Some of... [Read More]
Lessig Blog by Lawrence Lessig is licensed under a Creative Commons Attribution 3.0 United States License.
Comments (18)
Ownership, by customers or any entities,
is not a good idea.
It is better to add this clause to the
U.S. Constitution right after "To establish
Post Offices and post Roads;":
"To establish networks and endpoints;"
With that additional clause, the networks will
be truly commons without any kind of ownership,
just like the public domain. But then, the
chance of seeing networks being commons is
almost nil, especially in the light of what
the Congress did to the U.S. Postal Service.
Joseph Pietro Riolo
<riolo@voicenet.com>
Public domain notice: I put all of my expressions
in this comment in the public domain.
Very few people have a problem with the State owning and operating the roading network, providing a supplier-agnostic channel over which customers negotiate the delivery of services from a vast range of suppliers.
There is nothing inherent in the services that require owning the infrastructure... Maybe when telecommunications was a voltage in a wire there was, but the minute it was digitised you don't need to bind service to carriage. (Why am I telling you this when you often use the three layer model!?)
"Haven't we all learned that the market is more efficient at supplying goods and services?"
Yes we have, and we have learned from infrastructure monopolies that they are best owned by us all.
"Do we really need to rediscover the failings of Karl Marx at 100 megabits per second?"
100Mbs is not a service, its a delivery channel...
For a long standing municipal/private partnership fibre MAN, checkout http://CityLink.co.nz
And you still read http://Isen.com/blog?
Hamish.
PS. I've just discovered FA Hayek, "The Constitution of Liberty" Wow. What a vast argument for the freedom to tinker.
There is nothing inherent in the services that require owning the infrastructure…
when the answer is state ownership, the wrong question is usually being asked.
it's not the services, it's the infrastructure which is the issue. roads require land and a right of way, both of which are in limited supply. as a practical matter there can only be one owner of roads.
the "internet" is a completely different ballgame. you can get TCP/IP delivered to your door many differnet ways, ADSL, cable, wireless, etc. TCP/IP is anyway only one flavor. there could and should be a competing mix of proprietary and open formats carried on different infrastructure.
building a corresponding competing network of roads is simply impractical.
government ownership of "the internet" (either the infrastructure or the communication protocol which is carried over it) would be a very bad idea since it would interfere with what would/should otherwise be a competitive market for data services.
if there is a market for a "free and open" TCP/IP format, then consumers will buy bandwidth and support the cost of the infrastructure. (if not then the concept should be allowed to wither and die. the general public should never be taxed to provide this bandwidth.) if consumers prefer a content rich proprietary format then they should be able to buy that.
"TCP/IP is anyway only one flavor."
flavor of what ? and what besides IP comes to *your* house ? IPX ?
having many competing and nonstandard protocols is a bad idea.
The main problem with ethernet based networks is its asynchronous nature which is blindly driven from the data source and controlled by packet crash.
By allowing more, faster internet connections a very large amount of carrier free bandwidth is required. We could be headed for large patches of internet gridlock.
Shared ownership is good but how do we automatically build cost and profit sharing from major carriers down to small mum&dad investors/owners and end users into the whole network.
End users will generate increasing high bandwidth, low latency, low jitter, high cost traffic. The increased bandwidth also allows accelerated damage from trojans or hackers.
Access control has to come from the network with end users and routers regulated by the whole network itself. I'm in favour of a homogenous token/synchronous type network. Any type of protocols can be run end to end by the applications but the lower layers of hardware and routing will have to be controlled from within the network with a more wholistic view. The network terminators are a part of the 'whole' network reaching into the end users machine (telephone, pc, server, TV, ??)
Confluence
People keep saying, even here, that competition is good. Think about it people. Competition is good --- for consumers. It is bad for suppliers. When you leave capitalism alone long enough, it devolves down to an oligarchy. Most often it is called "consolidation" (remember when John Chambers at Cisco was buying 24 companies a year???), but it is really a drive by the suppliers to lessen competition (the 1890 Sherman antitrust act comes to mind). Think it isn't true? Then tell me how many baby bells we started with (only a few years ago when ATT was broken apart), and how many we have today.
When you get a market with a huge cost to entry like the "last mile" to your home, no one wants to supply you. The reason rural America has what phones and electricity it has is because the U.S. government required it in return for granting the monopolies. The Bell System, for example, used money from long distance and local phone tariffs to fund the wiring of rural America. They could not have done that in a competitive environment because of the lack of return on investment. They only agreed to do it under a regulated monopoly.
The same conditions exist today with broadband (if that's what you call something as achingly slow as DSL). What we need is three orders of magnitude faster than DSL (if you and your neighbors have any hope of seeing services like video on demand in your lifetimes). The only way we know to do that today is fiber. Fiber on the last mile.
No one is going to do this. No local exchange carrier, competitive or incumbent, is going to do this. Why? Because they don't have to. There is no competition on the last mile. None. The market is not going to do this because there are easier ways to get at least some return on their money.
The only way we are going to see fiber on the last mile is for government to build it and own it. That is, we the people must do it.
When we do this, we should have the wisdom to separate the fiber from the services that run on top of the fiber. The fiber should be open to all, just like the highways, for an access fee (payed by the suppliers of the services) that defrays the cost of running and maintaining the fiber.
If we don't do this, we get a situation where one of the baby bells in CA as I recall, was forced by law to open up their network to Covad. They had to supply Covad access to the last mile so that Covad could compete with them on DSL service. The rest, as they say, is history. Covad was forced into bankruptcy by the bell because the bell wouldn't move. It took years to gain access. Covad experienced delays in every conceivable way, especially if a line went down. Why? Because the Bell was also competing to provide the service, and it was in their enlightened best interest to work "to the book" for Covad.
The provider of the fiber must be forbidden from also providing a service that uses the fiber. Anything else is a conflict of interest.
If we don’t do this, we get a situation where one of the baby bells in CA as I recall, was forced by law to open up their network to Covad.
covad could have used cable, WiFi, 3G, satellite, etc. to connect you to their network. this is the point, there isn't only ONE road - insisting that all roads are equal and accessible to everyone diminishes the incentive to build new and different ones. "cyberspace" is not to be confused with real space.
“TCP/IP is anyway only one flavor.”
flavor of what ?
a communications protocol. like ethernet.
having many competing and nonstandard protocols is a bad idea.
yes, live would be simpler if everything was the same and everything was owned by the government, but it's already been tried.
Public ownership of infrastructure has several models that work. Roads. Electric utilities (there are around 2000 in the USA).
Over that last 100 years we discovered that creating a regulated body called a utility is beneficial in the last mile. That way the skies aren't lined with competing telecom carriers and the roads aren't ripped up by competing gas and water companies.
Yet today in telecommunications we have a situation where "the taxi company owns the road" (Brian K. Reid, PhD). If FedEd owned University Avenue, UPS and Emery would have no chance of making competitive deliveries in downtown Palo Alto. Similarly, how can one be surprised with the tepid results of DSL carriers like Covad and NorthPoint?
Muni Electric utilities have been operating a distribution network in the last mile successfully for 100 years. They buy power from multiple sources and distribute it in the last mile to their constituents. They are similarly well positioned to own and operate a Last Mile Network for telecommunications. They could operate a network much like a corporate IT department provides services to employees, and aggregate bandwidth in and out amongst several carriers, spreading cost and reliability risks. Any department (neighborhood or home) can pay for extra services to a carrier as long as network security and performance are not affected (herein lies a hint at open access).
Regarding the road/rail service/delivery content/carriage dichotomy, along with with concern about... CSMA/CA
http://www.bricklin.com/qos.htm
All hail the hypno-toad, but ATM didn't win.
Over that last 100 years we discovered that creating a regulated body called a utility is beneficial in the last mile.
only in the special case where LAND is involved. railroads, interstate highways, high tension electrical lines, telephone lines, gas, water, etc.
LAND is a limited resource - you can only build so many roads, pipelines, etc.
cyberspace is not so limited... it's like gauss said about infinity, it's more than a concept than a real number.
This is very interesting, good and informative page. I'm delighted with it! Thanks.
Hello. There are some factors with which I cannot to agree.
The interesting information located on your page gives I peep for mind more than a million audience to which I concern also.
I love the site! It's amazing, i visit every day and read your articles! Thanx very much
flavor of what ?
http://www.iglu.org/inet/
http://www.iglu.org/inet/
Common ownership would be great, but the RBOCs will sue at the drop of a hat to protect their monopolies. They have more attorneys and more money than most grass roots groups do, and municipality ownership is always under constant attack. To really be sustainable and free from legal interference by the RBOCs and political interfernce by lobbyists and political cronies, private sector ownership is the way to go. We've tried it three ways, and nothing else really worked for very long.