Waiting until the last minute as is my normal pattern, I have finally posted a comment to the Federal Communications Commission regarding Proceeding 14-28 titled “Protecting and Promoting the Open Internet”. Today is the deadline but I assume that everyone that actually has concerns about this topic knows that and is either suffering from procrastination like I have been or has already filed their comment, so this isn’t necessarily a call for you to get cracking and get your comment in too. On the other hand, if you haven’t yet planned to do this, you should: FCC Send Us Your Comments page
For the record, although my comments will appear on the FCC’s site in a day or so, here’s what I had to say:
Much of my life depends directly on an open and free-flowing internet. A part of this is that on a very small scale I am a web developer and hosting service. This represents a significant portion of my work and my income. In addition to that, I get virtually all of my news online. The vast majority of my communications come in and go out through the medium of e-mail. Although telephone calls to my immediate area are carried over copper pairs provided by Whidbey telecom, my local telco, most of my calls travel over the Internet using VOIP. I don’t watch television, I don’t subscribe to a dead trees newspaper, and the physical mail that comes to my post office box represents a tiny fraction of the incoming information to this household.
It is crucial for me, and I believe for the future of American society, that the Internet continues to serve all of its users, both publishers and consumers as well as merchants and customers, with scrupulous fairness in terms of transmission capacity and speed.
The business model that has made the Internet what it is today is relatively simple. On one side are the consumer facing ISPs who provide transmission to and from their customers, frequently local, who connect via telephone line, cable services, wireless or cellular connection, and doubtless a few creative niche alternatives that I’m forgetting at the moment. On the other side are the publishers of content, essentially anyone with a website, who either develop their own data centers and connections to the broader network or rent facilities from commercial hosts. Between these two sides exists what we call “the cloud”, the backbone carriers and all of the various interconnections and telco hotels that connect the consumer-facing ISPs on the one hand with the publishers on the other, both of whom charge their respective customers for their services, both of whom pay their upstream providers for connections to the cloud.
There has been one addition to this pattern, probably starting with the founding of Akamai in 1998, and that is what’s called the Content Delivery Network. This allows any publisher with a website hosted on a single server to engage the services of a CDN with globally dispersed proxies to speed up the experience of a user in Portland Oregon, for example, when browsing pages of a site located in Miami Florida or London England. Functionally this is an augmentation of the existing cloud. Clients of a CDN are using that system so that their sites appear to be nearby to users regardless of their physical location. Although the acceleration offered by a CDN directly benefits only the users of that system, by taking significant traffic off of the primary backbone providers every web browser in a specific location will tend to see content loading more quickly even on other sites. It’s hard to see this development as anything but positive.
Does such a system create a fast lane and a slow lane for Internet users? I believe the answer to that is clearly negative.
To take the question one step further, will the recent proposals from ISPs like Comcast to publishers like Netflix, in which those publishers pay a premium to those ISPs for faster delivery, create the dreaded two speed Internet? I have no direct knowledge of the mechanism by which this faster delivery will be accomplished, my assumption is that what is being offered is additional bandwidth across the cloud supplementing the current backbone providers. If this is the case, when Netflix pays extra to deliver their content more rapidly to Comcast data centers, all Comcast customers will benefit and no one will be relegated to the dreaded slow lane.
However, this is true only insofar as the ISPs continue to provide the same level of bandwidth between their data centers and the cloud that they were providing before the special service to those high-volume publishers was implemented. If Comcast and and Netflix contract to add a half gigabit per second of direct connectivity between Netflix and a specific Comcast data center, and Comcast maintains their existing cloud connections unchanged, no slow lane is created. In fact, customers viewing other sites will see an improvement in performance during those parts of the day when video traffic from Netflix is highest. However, if Comcast turns around and reduces their existing connections by the same half gigabit per second, customers viewing other content will see no immediate change in performance but over time, with growth in network volumes, their experience will begin to suffer.
Therefore the existence of these additional service contracts between publishers and ISPs has the potential to benefit the public, but only insofar as those ISPs do not reduce their current pattern of connections to the cloud which all their users will be exploiting in common. Whether ensuring that such contracts actually enhance network services requires a redefinition of the Commission’s structural approach to regulating Internet service providers or not is best left to the attorneys, I have no preference either way as long as the result of those contracts is an improvement for all users.
There is one question that has come up in the past that must not be forgotten: any attempt that an ISP might make to prioritize one type of traffic over another or to favor the traffic from one source over another, once that traffic has reached its data centers, must always be prohibited. This must apply regardless of the reason, whether an ISP simply doesn’t like BitTorrent, disapproves the politics of certain sites, or wants to block VOIP traffic to protect their own telephone business is irrelevant, there must be no favoritism shown by an ISP for one packet over another once that packet arrives at the datacenter.
If I Were King, net neutrality would be the law of the land with or without my comments.