I don’t think anyone seriously quarrels with bandwidth management. But it should be done in a protocol-neutral way. Those who use lots of bandwidth in a very short time should, if and only if network capacity is being reached, see their throughput curtailed until their demand drops off or the network congestion clears. But targeting a particular protocol, such as BitTorrent, is a completely different practice, and should not be tolerated. And Comcast didn’t just throttle BitTorrent — they rendered it unusable with spoofed “reset” packets that appeared to come from other BitTorrent machines.

Where things get sticky is when we start talking about preferential treatment. There is a good argument for giving priority to certain protocols whose nature is intolerant of delay — two-way video (not one-way, which can avoid network hiccups by buffering), real-time telemedicine (such as remote surgery, not sending an X-ray for interpretation, which can tolerate some delay), and the like. The question becomes whether such priority treatment should be restricted to particular protocols and whether it should be chargeable.

Generally speaking, I’m in favor of allowing the purchase of preferential treatment for purposes and protocols that aren’t delay-tolerant. But it is a very fine line that deserves careful walking. I also support traffic management for high-bandwidth users, as long as it is done solely to protect the network and other users’ experience on it, and ONLY when network loads demand it. But there is no excuse for targeting a protocol instead of a bandwidth hog. None.

NitrogenNick — You raise productive questions. I don’t know all the answers either, but I do think it would help if the industry developed standard protocols for network management (which I believe they’re attempting to do), clarified those protocols, published them, and consistently followed them. Transparancy and consistency would go a long way toward easing any remaining concerns.

The other factor that is not addressed by Pete’s article is the question of start up companies and how the Whitacres of telecom could strangle them in their cribs with large access fees if they thought it was in their interest to do so. The business model that those who oppose net neutrality want to implement could easily make the model that allowed for the creation of Google, Facebook, MySpace and YouTube impossible. This model depends on being able to have a good idea, good implementation and low start up costs. The elimination of net neutrality could kill this model.

It would say that it’s good that these debates are out in the open. The one who needs trust is the one responsible for demonstrating why he is trustworthy. It looks to me like the onus will fall on the net managers like Comcast.

*I also don’t know nearly enough about the specific mechanism to judge whether the accusations of deception have any merit.