Net Neutrality

The Internet its core technologies are based on packet switching equipment. The protocols (at least, the ones currently in use in the real world) and the equipment are not designed to provide a circuit that meets the requirements of multimedia applications.

Now, one logical area of Internet evolution is to provide all sorts of multimedia applications.

However, since the protocols and equipment aren't designed to provide circuits that meet the requirements of multimedia applications - this evolution is not going to occur without introducing new protocols and/or equipment.

One NECESSARY step in this direction is that the new protocols and/or equipment MUST be able to identify a multimedia application as it flows through the network.

Once it identifies that a particular flow is a multimedia application, it can reserve a multimedia circuit for this flow. Without this circuit, and using only today's equipment and protocols, the packets will not be delivered in a reliable and predictable sequence, making the multimedia experience for the end user become unacceptable.

Case in point - have you ever tried to watch streaming video over the Internet? It ain't pretty, is it? There's a long way to go before full screen video can be streamed over the Internet.

Notice in my argument I didn't mention anywhere about WHO was providing the multimedia content. I'm only talking about providing multimedia content - PERIOD. I don't care where it comes from, I only care that the technology can provide multimedia. Because, presently, it surely CANNOT provide multimedia - regardless as to who provides the content.

I claim Net Neutrality prohibits reliable multimedia from working on the Internet (I challenge anyone to explain HOW reliable, fullscreen multimedia streaming works in a Net Neutrality world)

One (flawed) argument might be to throw more bandwidth at the problem. If there's enough bandwidth, then surely any multimedia application will pass through the pipes.

This is flawed - simply because no multimedia service can be engineered on top of an unreliable delivery service (ie, the Internet) without building into it some sort of reliable, predictable delivery mechanism. Lots and lots of bandwidth addresses neither of those requirements - even with 99% bandwidth idle, a switch still might deliver packets in a non-predictable manner. This results in jitter, which, in sufficient quantity, renders a multimedia connection unusable.

How do we provide reliable, predictable, low-jitter connections? By reserving a circuit between two endpoints, with the characteristics of that circuit being sufficient for multimedia (reliable, predictable, low-jitter).

Once again, I'm not talking about where the content comes from - only that the content needs to get from point A to point B in a reliable, predictable fashion no matter who is providing the content.

However, I just described a process that is illegal in the Net Neutrality world. Setting up a reserved circuit is the same as distinguishing and discriminating one packet from another. Because as each packet enters a network, the equipment must decide whether it is destined for a reserved circuit or for general bandwidth. It gives preferrential treatment to those packets that are deemed part of a multimedia circuit - but that's not so much to crowd out other packets, but to simply guarantee that the packets are delivered in a reliable, predictable sequence.

So, if you want Net Neutrality, then you are saying you do not care to see multimedia applications on the Internet evolve beyond where they are today. Personally, I think you're excluding a lot of really cool stuff from being provided.