Joho the Blog
An Entry from the Archives

I was about to run an explanation by David Reed, one of the folks responsible for the neutrality of the Internet's architecture, of how the Internet works when Sen. Ted Stevens weighed in with his own. Stevens heads the committee writing the new telecommunications bill. So, I've decided to put them head to head in a death match. (If you've already seen the Senator's explanation, which I'm copying from Wired.com, you can click here to go straight to David Reed's.)

First, Senator Stevens. The fun part is working backwards from what he says to what his staffers told him. And beyond fun, we can gain insight into both what the Net discrimination side thinks and how it sounds to the, um, uninitiated:

There's one company now you can sign up and you can get a movie delivered to your house daily by delivery service. Okay. And currently it comes to your house, it gets put in the mail box when you get home and you change your order but you pay for that, right.

But this service isn't going to go through the interent and what you do is you just go to a place on the internet and you order your movie and guess what you can order ten of them delivered to you and the delivery charge is free.

Ten of them streaming across that internet and what happens to your own personal internet?

I just the other day got, an internet was sent by my staff at 10 o'clock in the morning on Friday and I just got it yesterday. Why?

Because it got tangled up with all these things going on the internet commercially.

So you want to talk about the consumer? Let's talk about you and me. We use this internet to communicate and we aren't using it for commercial purposes.

We aren't earning anything by going on that internet. Now I'm not saying you have to or you want to discrimnate against those people [...]

The regulatory approach is wrong. Your approach is regulatory in the sense that it says "No one can charge anyone for massively invading this world of the internet". No, I'm not finished. I want people to understand my position, I'm not going to take a lot of time. [?]

They want to deliver vast amounts of information over the internet. And again, the internet is not something you just dump something on. It's not a truck.

It's a series of tubes.

And if you don't understand those tubes can be filled and if they are filled, when you put your message in, it gets in line and its going to be delayed by anyone that puts into that tube enormous amounts of material, enormous amounts of material.

Now we have a separate Department of Defense internet now, did you know that?

Do you know why?

Because they have to have theirs delivered immediately. They can't afford getting delayed by other people.

[...]

Now I think these people are arguing whether they should be able to dump all that stuff on the internet ought to consider if they should develop a system themselves.

Maybe there is a place for a commercial net but it's not using what consumers use every day.

It's not using the messaging service that is essential to small businesses, to our operation of families.

The whole concept is that we should not go into this until someone shows that there is something that has been done that really is a viloation of net neutraility that hits you and me.

The full audio is here.

Now for David Reed. This comes from an email list, with David's permission. The list is discussing the DPS Project Net neutrality bill David favors. (Me too.) I've added in some links and have lopped off some intro and extro material. David is writing in response to someone who says that the DPS Project bill would stifle innovation.

Why is the Internet different from PSTN [public switched telephone network], cable, or broadcast TV?

The secret is that the Internet architecture was designed so that the transport network itself does not need to be involved in innovations. This is due to following the so-called "end-to-end argument" that I helped describe and name, but which was developed for the specific purpose of leaving the possibilities for innovation as open as possible. I do take credit, not for being old or gray in beard, but for being part of the group who made the Internet work this way.

As Dave Clark has said, the key to the Internet's ability to absorb and to benefit from innovations for the last 30 years without breaking is the so-called "hourglass" architecture, where the thin neck is at the point where the IP protocol sits. Innovations in technology are deployed either in the layers below IP (which do not affect applicaitons - so that such ideas as all-optical switching, ad hoc wireless meshes, etc. need not change the definition of the Internet one whit) or in the layers above IP (so that ideas such as VoIP, WWW, Internet mail, chat rooms, social networking or Wikis require no change to the Internet to be tried and adopted).

So the only thing we must consider is the IP protocol, which is quite simple and amenable to many mappings.

What the [DPS Project] bill does at its core is require anyone who provides "access to the Internet" to provide access to the consensually defined Internet as of a particular point in time.

What does that mean? Well, it's quite simple, actually. The Internet merely provides a service such that one has an IP address, and that from that IP address, one can send a message (technically called an IP datagram) to any other IP address. The "envelope" of that message contains a set of fields that specify the source, the destination, and certain handling features such as a "protocol number". The message itself is a string of bits that are utterly meaningless to any device other than the source or the destination - in fact, the message can be strongly encrypted with a key that is known only to the source and the destination, or formatted in any way that makes sense to the endpoints.

So exactly what kind of innovations would be prevented in the Internet? Well, there is one kind that would be prevented. That is an architecture that requires devices between the source and the destination to decode and understand the messages being sent.

That would be an innovation, but it would be the *last* innovation that ever happens to the Internet, for a very simple reason. Once the network starts to require understanding of the communications that used to be merely the business of the source and the destination, new innovations of services would require permission and redesign of the Internet every time a new service were introduced.

That is why we eschewed putting such processing in the network, and invented the end-to-end arguments that led to the Internet design. We were facilitating a state of constant revolution and change, both in the applications that could be deployed and in the techniques used to move bits from one place to another.

It is this state of constant revolution and change that I am proud to have helped enable. It is for this reason that conservatism in preserving the core architectural principle of the Internet is warranted precisely because it facilitates change and innovation, both above and below the IP layer.

Innovations that don't refer to a product or service labeled as "access to the Internet" would of course be unaffected [by the DSP Project bill].