Vint Cerf can.
It’s ironic that the man known as the “Father of the Internet” is now best-positioned to do what is needed to save it. To save it as a neutral medium, an end-to-end “network of networks.”
Cerf is currently “chief evangelist” at Google. He joined it from MCI, now part of Verizon. Mostly, he’s a spokesman, a Washington-based attender of cocktail parties, the old guy they trot out and hand plaques to when they want Capitol street cred. “Here’s the Father of the Internet. He works for us.”
But assuming net neutrality loses in the Senate, Google faces a problem. The Bells are determined to cut it off at the knees, either demanding blackmail (they’ll call it “congestion fees” but it’s the same thing) or cutting deals with rivals (Yahoo) to slow access to Google services.
And many of the services Google is bringing on-stream need a fast, neutral net to work. Google Talk. Video search.
But there is a way out.
Dark fiber.
Google owns tons of dark fiber. This is fiber cable that’s underground and, on the whole, unlit. I have no doubt that Google uses some of it to connect its own operations, to save a few bucks.
I don’t know how much Google has, but I’m guessing it’s enough to connect every major market in the U.S., and (with a little investment) most of the minor ones. If I’m wrong, buying a company like Level 3 (quite affordable) solves the problem. (It also brings a big team in. Maybe even a stake in Level 3 would get the job done.)
What to do with it?
Offer connections via this fiber to the Internet backbone for anyone who wants them, at rock-bottom prices. Work with Intel and other vendors to deliver WiFi and WiMax gear that can reach these points with as much speed as possible.
Let a thousand ISPs bloom. Become the Costco of backhaul. And remember, the more traffic you’re handling the more the Bells have to deal with you as an equal on peering as well.
Suddenly folks will find they have choices other than the Bell duopoly. Corporations, schools, and others who need backhaul will find a supplier offering rock-bottom prices. The Bells will also have competition in the core (they now own both AT&T and MCI) so they will be unable to raise prices there.
They will be beaten.
But for this to happen, Cerf needs to have the vision, the energy, and the budget to fight for it. He has to get Sergey and Larry on his side. He has to get it through the board. And he has to put together a technical team to get this job done.
It can be done. Someone has to do it. It’s low-margin, yes, but it will do more than anything to boost the entire US Internet sector, which needs boosting.
Please, Vint. Do it for the children.
Let me relate a story from the past to reinforce what Dana is saying in this post.
Many years ago, when MCI was independent and still a young company, they entered the consumer long distance business. At a meeting with MCI executives, I asked why they did this. They said that their primary business was selling telecommunication services to business. The move into the consumer market was defensive.
At the time, ATT owned nearly 100% of the consumer long distance market. MCI’s fear was that ATT would use its monopoly power in consumer long distance to drive up margins in the consumer market. Higher consumer long distance margins would then be used subsidize an attack on MCI’s business customer base.
As a purely defensive move MCI reasoned that if they took 10% of the consumer long distance market they could keep margins from rising and as a result rob ATT of the financial resources to launch an attack in the business market.
I intrepret Dana’s post to be saying that Google and other national service providers would be well served to ensure that the Internet backbone business remains a commodity market. As with MCI, it is a wise defensive strategy.
That said, you don’t win playing defense alone. You need an offensive strategy to go with your defensive strategy.
When MCI entered the consumer long distance business, local access telephone networks were common carriers. The local telephone company could not discriminate between long distance service providers. The local access networks were not owned by ATT.
Today, the telephone companies own major Internet backbones and the access networks. They are not common carriers. They are free to discriminate among service providers that want to use their networks. Worse, they are free to discriminate against independent service providers and in favor of the services they sell.
In the new world, Internet access service is just another service the telephone companies sell. It is their service. They can limit the amount of bandwidth and regulate the price of Internet service. By throttling Internet access service, they can favor with more and cheaper bandwidth, services sold in their walled access gardens.
Net neutrality does not prevent the telephone companies from starving Internet service for bandwidth. It just makes sure the telephone companies cannot discriminate among Internet services we sip through the straw they provide us. We are on a course where high bandwidth service innovation comes under the control of the telephone companies.
The effect will be to force new and innovative high bandwith services into the telephone company walled gardens where margin can be transferred from service providers like Google to the telephone companies.
There are three offensive strategies available to service companies like Google. First, is to do nothing. Let the telephone companies take their margin.
Second, is to fight in Congress for telephone companies to be transformed back into common carriers. The problem is that this will take a long time and the results uncertain. Service providers do not control the ground on which this battle will be fought.
The third strategy is to back third parties building competitive IP common carrier access networks.
Wireless Internet access can provide some competition to the telephone companies. It puts marginal pressure on telephone company Internet service margins.
But, wireless Internet service is not enough to break the telephone companies stranglehold over the future of the Internet. The telephone companies are still in a position to transfer margin from the service providers to themselves for new service that require high bandwidth.
The best long term strategy is for Google and the other service providers to support the construction of common carrier IP access networks built from fiber. This strategy keeps the telephone companies from gaining a monopoly over the high bandwidth services market.
Tactically, there are two immediate actions service providers like Google must take. First, is to trade a win on net neutrality for perserving the right of municipal governments to build common carrier IP access networks. These networks will give Internet-based service providers the ability to compete with telephone and cable company services sold in walled gardens.
Second, is for Google and other service providers to provide intellectual support to the municipal governments wanting to build common carrier IP networks (wireless/fiber).
Short term, Google and others interested in promoting common carrier IP access networks should build a prototype so the general public can experience its possibilities. Make this a traveling exposition that goes from community to community. Give the public an opportunity to experience what it is missing.
In summary, Google and other service providers that depend on an open Internet must do three things.
1)Ensure competition in the Internet backbone service business.
2)Ensure that federal legislation is not passed that prevents municipal governments from building common carrier IP networks.
3) Invest intellectural capital to create a model of a working municipal common carrier IP network for traveling exposition. The goal of the traveling exposition is to educate the general public by giving them the opportunity to experience the services such a network makes possible.
Let me relate a story from the past to reinforce what Dana is saying in this post.
Many years ago, when MCI was independent and still a young company, they entered the consumer long distance business. At a meeting with MCI executives, I asked why they did this. They said that their primary business was selling telecommunication services to business. The move into the consumer market was defensive.
At the time, ATT owned nearly 100% of the consumer long distance market. MCI’s fear was that ATT would use its monopoly power in consumer long distance to drive up margins in the consumer market. Higher consumer long distance margins would then be used subsidize an attack on MCI’s business customer base.
As a purely defensive move MCI reasoned that if they took 10% of the consumer long distance market they could keep margins from rising and as a result rob ATT of the financial resources to launch an attack in the business market.
I intrepret Dana’s post to be saying that Google and other national service providers would be well served to ensure that the Internet backbone business remains a commodity market. As with MCI, it is a wise defensive strategy.
That said, you don’t win playing defense alone. You need an offensive strategy to go with your defensive strategy.
When MCI entered the consumer long distance business, local access telephone networks were common carriers. The local telephone company could not discriminate between long distance service providers. The local access networks were not owned by ATT.
Today, the telephone companies own major Internet backbones and the access networks. They are not common carriers. They are free to discriminate among service providers that want to use their networks. Worse, they are free to discriminate against independent service providers and in favor of the services they sell.
In the new world, Internet access service is just another service the telephone companies sell. It is their service. They can limit the amount of bandwidth and regulate the price of Internet service. By throttling Internet access service, they can favor with more and cheaper bandwidth, services sold in their walled access gardens.
Net neutrality does not prevent the telephone companies from starving Internet service for bandwidth. It just makes sure the telephone companies cannot discriminate among Internet services we sip through the straw they provide us. We are on a course where high bandwidth service innovation comes under the control of the telephone companies.
The effect will be to force new and innovative high bandwith services into the telephone company walled gardens where margin can be transferred from service providers like Google to the telephone companies.
There are three offensive strategies available to service companies like Google. First, is to do nothing. Let the telephone companies take their margin.
Second, is to fight in Congress for telephone companies to be transformed back into common carriers. The problem is that this will take a long time and the results uncertain. Service providers do not control the ground on which this battle will be fought.
The third strategy is to back third parties building competitive IP common carrier access networks.
Wireless Internet access can provide some competition to the telephone companies. It puts marginal pressure on telephone company Internet service margins.
But, wireless Internet service is not enough to break the telephone companies stranglehold over the future of the Internet. The telephone companies are still in a position to transfer margin from the service providers to themselves for new service that require high bandwidth.
The best long term strategy is for Google and the other service providers to support the construction of common carrier IP access networks built from fiber. This strategy keeps the telephone companies from gaining a monopoly over the high bandwidth services market.
Tactically, there are two immediate actions service providers like Google must take. First, is to trade a win on net neutrality for perserving the right of municipal governments to build common carrier IP access networks. These networks will give Internet-based service providers the ability to compete with telephone and cable company services sold in walled gardens.
Second, is for Google and other service providers to provide intellectual support to the municipal governments wanting to build common carrier IP networks (wireless/fiber).
Short term, Google and others interested in promoting common carrier IP access networks should build a prototype so the general public can experience its possibilities. Make this a traveling exposition that goes from community to community. Give the public an opportunity to experience what it is missing.
In summary, Google and other service providers that depend on an open Internet must do three things.
1)Ensure competition in the Internet backbone service business.
2)Ensure that federal legislation is not passed that prevents municipal governments from building common carrier IP networks.
3) Invest intellectural capital to create a model of a working municipal common carrier IP network for traveling exposition. The goal of the traveling exposition is to educate the general public by giving them the opportunity to experience the services such a network makes possible.
I think you guys need to take a deep breath and calm down before you go into cardiac arrest.
Charging a fee for QoS isn’t the End of the Internet. Commercial Internet customers have been able to buy QoS-based service plans for several years and the Daily Kos is doing just fine. They even had a convention on Las Vegas this weekend.
The Internet backbone is already extremely competitive, and has been forever, without any regulations on it at all. If only the consumer Internet were that competitive.
Google will never invest a dime in consumer hookups because they benefit much more from using the dark fiber to run a private network with portals near every major population center. They do this today.
Net neutrality regulations that forbid for-fee QoS make the ISPs Google’s slaves. This deal is so sweet for them it’s as if FedEx could stuff their letters in the USPS’s mail pouches and have them delivered same day for the cost of bulk mail.
Don’t be so gullible. This is not a good-guys vs. bad guys issue. It’s a complicated telecom and network engineering issue that has to be dealt with soberly and analytically.
All this huffing and puffing isn’t helping, and I doubt it even makes you feel good.
I think you guys need to take a deep breath and calm down before you go into cardiac arrest.
Charging a fee for QoS isn’t the End of the Internet. Commercial Internet customers have been able to buy QoS-based service plans for several years and the Daily Kos is doing just fine. They even had a convention on Las Vegas this weekend.
The Internet backbone is already extremely competitive, and has been forever, without any regulations on it at all. If only the consumer Internet were that competitive.
Google will never invest a dime in consumer hookups because they benefit much more from using the dark fiber to run a private network with portals near every major population center. They do this today.
Net neutrality regulations that forbid for-fee QoS make the ISPs Google’s slaves. This deal is so sweet for them it’s as if FedEx could stuff their letters in the USPS’s mail pouches and have them delivered same day for the cost of bulk mail.
Don’t be so gullible. This is not a good-guys vs. bad guys issue. It’s a complicated telecom and network engineering issue that has to be dealt with soberly and analytically.
All this huffing and puffing isn’t helping, and I doubt it even makes you feel good.
Richard, don’t be so gullible yourself. Net Neutrality has nothing to do with QOS. It has to do common carriage. Do the network operators get to price people who might compete with their own services out of the market or do things remain as they are today where you pay for traffic based solely on volume not type? Their is nothing wrong with tiering, per se, but the process through which it is implemented needs to stay transparent. Physical networks are always going to be a natural monopoly and that means they need to be regulated. To take the blanket view that all regulation is inherently bad is religion, not reason.
Richard, don’t be so gullible yourself. Net Neutrality has nothing to do with QOS. It has to do common carriage. Do the network operators get to price people who might compete with their own services out of the market or do things remain as they are today where you pay for traffic based solely on volume not type? Their is nothing wrong with tiering, per se, but the process through which it is implemented needs to stay transparent. Physical networks are always going to be a natural monopoly and that means they need to be regulated. To take the blanket view that all regulation is inherently bad is religion, not reason.
No no no, Jesse, it has everything to do with QoS. I suggest you do something that few neutrality regulationists have ever done: read the actual bill. The Markey Amendment says:
If a broadband network provider prioritizes or offers enhanced quality of service to data of a particular type, it must prioritize or offer enhanced quality of service to all data of that type (regardless of the origin or ownership of such data) without imposing a surcharge or other consideration for such prioritization or enhanced quality of service.
And similar language exists in Snowe-Dorgan and Sensenbrenner-Conyers.
The people who are screaming the loudest about net neutrality are concerned about free speech, blogs being blocked and that sort of thing. But the actual regulations they propose have nothing to do with speech and everything to do with protecting Google’s business position.
What’s the connection between free speech and QoS? As you say, really nothing.
So why ban for-fee QoS?
No no no, Jesse, it has everything to do with QoS. I suggest you do something that few neutrality regulationists have ever done: read the actual bill. The Markey Amendment says:
If a broadband network provider prioritizes or offers enhanced quality of service to data of a particular type, it must prioritize or offer enhanced quality of service to all data of that type (regardless of the origin or ownership of such data) without imposing a surcharge or other consideration for such prioritization or enhanced quality of service.
And similar language exists in Snowe-Dorgan and Sensenbrenner-Conyers.
The people who are screaming the loudest about net neutrality are concerned about free speech, blogs being blocked and that sort of thing. But the actual regulations they propose have nothing to do with speech and everything to do with protecting Google’s business position.
What’s the connection between free speech and QoS? As you say, really nothing.
So why ban for-fee QoS?
By the way, here’s my favorite Vint Cerf quote:
Internet is for everyone – but it won’t be if Governments restrict access to it, so we must dedicate ourselves to keeping the network unrestricted, unfettered and unregulated. RFC 3271, http://tools.ietf.org/html/3271
He was working for the Phone Company when he said that.
By the way, here’s my favorite Vint Cerf quote:
Internet is for everyone – but it won’t be if Governments restrict access to it, so we must dedicate ourselves to keeping the network unrestricted, unfettered and unregulated. RFC 3271, http://tools.ietf.org/html/3271
He was working for the Phone Company when he said that.
How can you really sell QoS on a connection-free network, one where you can’t guarantee that you control the bits front-to-back?
You can’t.
The whole QoS argument, thus, is based on the idea of a monopoly provider. It’s contrary to the Internet’s design. It’s a straw man.
How can you really sell QoS on a connection-free network, one where you can’t guarantee that you control the bits front-to-back?
You can’t.
The whole QoS argument, thus, is based on the idea of a monopoly provider. It’s contrary to the Internet’s design. It’s a straw man.
It’s actually not hard to superimpose connections on a datagram network: that’s what TCP does.
So don’t worry about how to do it, it’s all laid out in the RFCs for DiffServ and SIP.
And anyhow, if this is impossible, why do the Markey and Snowe-Dorgan laws try to ban it? That would be a waste of time if what you’re saying were true.
It’s actually not hard to superimpose connections on a datagram network: that’s what TCP does.
So don’t worry about how to do it, it’s all laid out in the RFCs for DiffServ and SIP.
And anyhow, if this is impossible, why do the Markey and Snowe-Dorgan laws try to ban it? That would be a waste of time if what you’re saying were true.