This is the full, uncut copy I submitted to the Mercury News. The published piece (In neutrality debate, carriers get blamed for Net’s weaknesses) has moved to the Merc’s archives, where you have to pay to retrieve it.
The Circus is Coming
The circus is coming to Palo Alto. The FCC’s network neutrality circus that is, the dramatic battle between two conflicting views of the Internet. In this tussle, the lovely but fanciful notion of a semi-divine and nearly perfect engine of democracy and community sets itself against the reality that today’s Internet is a warty gadget that lives on the edge of collapse in the best of times.
The FCC is investigating a group of complaints from the consumer protection lobby and a local startup, Vuze, Inc., against Silicon Valley’s cable company, Comcast. The complaints allege seven different kinds of villainy and seek enormous fines. The Commission has already held one public hearing, sponsored by the Berkman Center for Internet and Society at the Harvard Law School, and holds the follow-up at Berkman Center alumnus Larry Lessig’s Stanford Center for Internet and Society.
Little good came from the Berkman hearing. Both panels were dominated by legal scholars, academics, and business interests dead set on securing free rides. Vuze was given its own time and the special privilege of a multi-media presentation, while ordinary witnesses encountered resistance from the Commission in simply showing Power Point slides, let alone short video clips (such as the Web Hog commercial from 2000, that I wanted to show.)
Chairman Martin made no secret of his sympathies. He badgered Comcast’s solitary witness after fairly swooning over Vuze and failed to display the slightest insight into the management challenges faced by broadband carriers. The Internet was designed for the polite society of network engineering professors and their graduate students, not our rough-and-tumble world of viruses, e-mail scams, and copyright theft, and it shows. Peer-to-peer applications, such as the open source version of BitTorrent used by Vuze, are designed to consume a disproportionate share of network bandwidth, and carriers have to limit this appetite to provide good service to mainstream users. Japan has learned that adding more capacity to the network doesn’t alleviate this problem: peer-to-peer consumes the largest share of the pipe, no matter how big it is.
It should come as no surprise that the Internet’s official doctrine fails to provide necessary tools or that popular implementations fail to follow best practices. The Internet has always been vulnerable to the traffic demands of new applications: researchers encountered â€œInternet meltdownâ€ in the mid-80s, and new users encountered a â€œWorld-Wide Waitâ€ in the mid-90s. The solution to these problems was re-engineering of vital protocols, an increasingly dicey proposition. The Internet’s much vaunted â€œend-to-end architectureâ€ means that upgrades to vital systems roll out at a glacial pace measured in years and even decades rather than the â€œInternet timeâ€ we heard so much about during the last bubble. The latest advance in Internet traffic management, Explicit Congestion Notification, was standardized in 2001 and remains dormant in Windows Vista.
Against this background broadband carriers struggle to balance of cost, performance, and fairness, all the while hectored by a consumer rights lobby oblivious to their technical challenges.
The public is unlikely to benefit from the FCC’s protracted hearing process unless there’s a change of emphasis. Comcast has already announced a change to their traffic management system that makes it more application-agnostic, so the basis of the complaints will soon be moot.
The Commission should use the Stanford hearing to explore the ways that the Internet is unique among networks, and to improving the means available to consumers for evaluating competing Internet access services. For openers, this begins with the understanding that residential Internet access is a variable-rate system. These services are sold with caps on download and upload rates, not with guarantees that the peak rate is available at all hours (commercial Internet accounts with guaranteed â€œCommitted Information Ratesâ€ are much more expensive than residential accounts.)
It would be helpful to consumers to know what rates they can expect, on average, at different times of the day, and to have additional management capabilities. End-user traffic shaping is an especially important feature for those of us with home networks.
While the FCC can help in getting a disclosure process moving, it’s not necessary to force it by creating new rules for network management. Industry, in collaboration with responsible consumer groups, is already headed in that direction, so it’s sufficient for the Commission to take a secondary role. The best thing it can do is lead by the example of conducting a more sober hearing than the Berkman Affair.