Blog Post

Another Perspective on Net Neutrality: The Kids Are Alright

Yesterday, Federal Communications Commission Chairman Julius Genachowski unveiled a framework for the regulatory agency’s net neutrality efforts — the idea that a broadband provider cannot interference with the delivery of lawful bits over its network — that received fairly lackluster support across the board. While my immediate coverage of the framework focused primarily on the fact that Genachowski opened the door to usage-based billing, others spent a lot of time dissecting the various loopholes that could allow operators and ISPs the ability to actively interfere with the web experience. But David P. Reed, a developer of several networking technologies core to the Internet, provides a really compelling counterpoint to all the moaning and groaning, by explaining exactly how the FCC has gotten things right so far.

In a blog post today, Reed discusses how Genachowski has so far been careful to distinguish between the Internet itself and how people access the Internet — namely, broadband. That same issue, the transfer of the bits rather than the actual bits themselves, is at the heart of the whole reclassification debate that could determine how much power the FCC has to “regulate the Internet.”

At issue is if broadband is a transport service or an information service. The FCC has the power to regulate transport, but for now, may not be able to regulate “broadband” thanks to a series of regulations passed by previous FCCs that classified broadband as an information service. Whether or not this changes has large implications for the FCC’s abilities to remain relevant in a broadband world, but legal arguments aside, Reed writes that by making a distinction between broadband as an access technology and the Internet as a destination Genachowski got something right:

The key distinction here is that letting the users, through their applications, choose what their packets do, and where their packets go, and specify whatever priorities they need is quite different from having one link in the chain, the last mile provider, be able to control what applications the user can effectively run, or constrain how well they run. It’s this edge based freedom to choose to whom you speak, with what services you connect, how you speak and what you can collaborate on that constitutes the freedom of the Internet, due to its openness. The last mile access provider (Verizon, ATT, Cogent on the consumer side, or various large interconnects, such as Level 3, that support business users like eBay or small businesses) should have no say in deciding what Internet applications do inside their packets.

The main risk here could be that a provider makes something that they call the Internet, but in fact is highly non-standard and is not at all a standard Internet connection. That would be clear interference with the Internet, and if folks are required to be transparent, their deviations will be apparent, and the relief can be sought (either by the user or by appeal to the regulator). By distinguishing the Internet from other Broadband offerings, we can at least clarify the discussion, and if there are specific guidelines keeping the Internet open, by making sure it is the same Internet all over the US and the world, then we have a basis for sensibly understanding what will destroy its value.

The key benefit of framing things this way is that the Internet connection sold to the user must provide the standard, commonly understood, consensus definition of the Internet service – the ability to address IP datagrams to any destination on the collective Internet, and have it delivered without the access provider interfering with that delivery for its own purposes, or doing something other than deliver the datagram as requested according to the standard understanding of how the Internet works.

I appreciate the distinction Reed is drawing here, but I’m not sure all the careful definition setting will matter if the lawyers at the FCC can’t uphold and defend their right to enforce network neutrality, or if those loopholes that Reed acknowledges enable ISPs to make it difficult, expensive or unpleasant for consumers to access the true Internet.

Image courtesy of Gavin St. Ours on Flickr.

Related GigaOM Pro Content (sub req’d):

One Response to “Another Perspective on Net Neutrality: The Kids Are Alright”

  1. I tried to leave a comment on Reed’s blog yesterday, and he’s censored it. So here it is:

    “I appreciate your desire for linguistic exactitude, David. When an engineer uses a term, it has a much narrower scope than it does when a policy advocate uses the same term. The scope differences in such terms as “discrimination” has kept the debate over Internet regulation alive far beyond its useful life.

    “However, I find a great deal of inexactitude in your post once you get off the broadband vs. Internet point. You say, for example: “The Internet was designed to work over any kind of network, and it works over mobile broadband platforms today!”

    “What do you mean by “was designed?” Do you mean there was a hope, or a goal, or an intention to make the Internet work over networks that didn’t exist in 1974, or that everything needed to ensure it would work was done in the beginning?

    “And what do you mean by “work?” Do you mean “stumble along, losing every other packet and sucking down an enormous electrical load but more or less connecting everyone to everyone as long as nobody’s in a hurry” or do you mean “functions efficiently across the broadest conceivable range of applications and networks, promotes innovation and investment, and costs next to nothing to operate?”

    “These terms are also important elements of the debate over broadband and Internet regulation, and need to be precisely defined as well.”

    David Reed is barking up the wrong tree. The debate over Internet regulation is not a question of hair-splitting and semantics, it’s a policy debate about who regulates the Internet, how much power the regulator has, and in what direction we want the regulator to push it. These questions fundamentally need to be resolved by Congress, and there’s much more at stake than simple definitions.

    Incidentally, Reed is acknowledged to have played a part in the decision to separate TCP from IP in 1978, and has played no discernible role in the development of networking technology since then, so you’ve overstated his credentials just a bit.