> Thanks for the thoughtful response. Care to explain your CDN/ISP hypothetical further? Wouldn't the content providers who are capable of paying have an advantage over those who aren't?
Of course, just as Amazon has an advantage over smaller retailers, and Netflix has an advantage over new steaming services. Incumbents always have certain advantages, and business deals always favor those involved in them over those not involved in them. Money makes things easier. But "an advantage" isn't "an insurmountable advantage", and HN is full of startups coming up with innovative ways to circumvent incumbent advantages.
> I'm not clear on how money changes hands in your case, but undoubtedly this arrangement would favor the ISP.
Not necessarily. This doesn't need to be a win/lose zero-sum scenario; it's entirely possible, in the scenario I described, for all three of the customer, the ISP, and the content provider to come out ahead.
> As far as peering goes, it's a bit more complicated... server locations are more mobile than consumer locations, so there is less concern about an effective geographic monopoly for backbone providers.
That's certainly true. And in general, I think it's much more critical to find a solution to the ISP monopoly problem than to the network neutrality problem, because ISP monopolies lead to quite a few other issues as well.
> Without fully thinking through the economics of peering, my initial thought is that there should probably be a regulated amount that ISP's can charge to peer with backbone providers. Maybe cost plus a maintenance fee based on traffic volume.
I can understand where that idea would come from, but ick; there are so many things wrong with that idea.
Perhaps it would make sense to legally separate the "last-mile" provider from the large-scale bandwidth provider, with the former providing only connectivity to a local meet-me room. I don't particularly like that idea, but it seems far less awful than alternatives.
