Monday, March 31, 2008

Paying Per Packet? Don't Be Selfish

Brad Templeton serves as chairman of the Electronic Freedom Foundation and is a blogger. What follows is a near-realtime transcript of his remarks. If you have any additions or amendments, let me know.

I am the chairman of the EFF. It's not saving the Internet that's hard, it's audio visual. We got into a battle in this town against the president of the United States, suing AT&T for letting everyone wiretap. I want to talk about the history of regulation in data communications. I'm also involved in BitTorrent Corp., but I'm not speaking officially on behalf of that today.

Most of the people here are in favor of an open network. The Net was created as a peer-to-peer environment. This allowed people to come up with the concept of the end-to-end or stupid network. It's a great thing, and hopefully we have a choir here.

But the real invention of the Internet wasn't a technological invention. Believe it or not, the thing that made the Internet great was it's pricing model. It was an economic invention. It's pricing model didn't involve money. I pay for my line in the middle, you pay for your line to the middle, and we don't care about what happens in the middle. We don't sweat the small stuff. That still involves paying for things but enabled a whole bunch of applications.

Paying per packet means that every application on the network had to be justifiable. Network providers are run by bean counters. One of the earliest Web sites was a video camera trained on a fish tank. That would have been shut down by the network. But some of the best applications aren't financially justifiable. Paying per packet also leads to a network of timid users. The cost of books shouldn't matter to us. But how many of you see a book that's $1 more than it should be? Paying money on an incremental basis has a psychological cost.

There's a bit of a monster hiding in the pricing model. Everybody oversells their Internet capacity now. They really want to sell you shared access to a pipe they don't intend you to use. They sell you oversaturated pipes. They advertise unlimited use. Is that a business? Whose bandwidth is it? Do they sell the bandwidth to the customer, or don't they? There's a dissonance.

Meanwhile, the upstream component goes unused most of the time. That got exploited by peer-to-peer publishing tools. Peer-to-peer file distribution is the best technology for publishing a file cheaper. It's no surprise that people who want to do copyright infringement pick the best technology to do it. It's used for both honest distribution and otherwise.

There's always going to be some application that's a bandwidth hog. There's always going to be one that uses more bandwidth versus less bandwidth. We don't want to beat down the winner. Right now, peer-to-peer is the villain, but if you get rid of that, there will be another one.

I'm going to go in the opposite direction when it comes up with coming up with a law for Net neutrality. I'm going to make a rather bold statement that's not entirely true. All the telecom regulations have actually caused more harm than good. They all turn sour in the long run. They do some good in the early days, but before long, they go bad.

Simply having paperwork and asking raises an effective barrier. Even if it seems like you're not asking for too much. Solving Net neutrality is like putting out a fire with corn-based ethanol, which is itself the result of activity by the corn lobby. We've got the telcos in a briar patch. Br'er telco is more nimble in the briar patch than br'er fox. We're br'er fox.

It subsidizes phone service for poor people. Long ago, maybe that helped. Years ago, a friend and I decided we'd drop a phone booth in the middle of the Black Rock Desert. You can actually bring wireless to the rural user cheaper than you can put in an urban landline. Two guys could do that. If you need to subsidize something, just subsidize it.

E911 is another example. If you want help in an emergency, you don't want to say that something that looks like a phone is the means to it. $1/use/month kills… Let's look at CALEA. Why didn't we learn more from 80211?

If I were put in charge, I would replace the FCC with three words, "Don't be selfish." Don't interfere with others. Don't be overly sensitive to interference.

The one regulation I would say is somewhat successful is the one we're suing AT&T over. Back in the '70s, there was a president that Congress decided wan't trustable. He did wiretaps. So they made a law that said phone companies couldn't allow wiretaps on all traffic without warrants. We sued them under that law. So the president got the Senate to pass a law. The House hasn't passed it. Thank you, House.

Technology changes so quickly in telecommunications, that any policy should automatically sunset. The time they change from good to harm is remarkably short. Be careful what policies you have. Review all policies every few years. It's the monopoly, stupid! There's no need for a natural monopoly. We've created these monopolies and allowed people to use them. We need to get in there and use the dark fiber.

There's plenty of bandwidth. If we really open this up, I'd like to see people able to go into small electronics stores to install small networks in their neighborhoods by just burying the fiber in their yards. Make a technology mass market, and lo and behold, it's $20 at Fry's in just a few years. This technology is possible.

People say the Internet can't scale for video. That's totally wrong. IP multicasting can work if we want it. Peer-to-peer also scales up. It can provide the ability to distribute vast amounts of data. What's popular only goes over the line one or two times, and then it's in the local cloud. A lot of people who built these networks made the assumption that everybody would be a consumer.

Comcast did a deal with BitTorrent on Thursday or Friday. What Comcast did was sort of sly. They didn't do that much interference with BitTorrent. They put detectors on the network to see what people had finished downloading. Then you become a seed, and other people can access it. If you were seeding, and if you were seeding people who weren't on Comcast, that's when they went in and decided they wanted to stop you. The thing that got them in trouble was putting in synthetic packets that would say that the person on the other end had hung up.

The network rose up about that. There was regulatory threat. Comcast realized that, crap, we can't even do this. They've agreed now that they're going to be protocol agnostic. They're going to balance bandwidth. They'll put more upstream in the network. They'll do new research so there's more local caching. And they're going to be more transparent about what they do.

The best lessons in the network come from the edge, both in technology and network neutrality.

No comments: