I have always enjoyed Amazon's forays into net neutrality - and since 2006, they have argued for negative net neutrality, as have I. Now Paul Misener argue for FRAND - allcomers can access higher QoS so long as that does not slow regular traffic. How does that work? Simples - increased capacity to match that new QoS traffic. Its exactly what I described in the book, and if it is to happen, its not just Amazon who will celebrate.
"Why should one Web site be able to pay for better performance when others do not or cannot pay? The answer is simple: the improved performance of this Web site has not come at the expense of any other Web sites, and the same enhancements were available to everyone else. One site paid for performance enhancement, but others didn't suffer degradation. If paid performance enhancement for some content is equally available and does not degrade the performance of other content, then it should be permissible. And, following this principle, in addition to moving, leasing private lines, and edge caching, Internet content providers (and consumers) should be able to purchase "quality of service" or "managed services" from network operators on the same basis--equal availability and no harm to other content. At a recent conference, this approach to Net neutrality was described, and another speaker remarked that, at a completely full network bottleneck, it is impossible to favor some content without degrading other content. This is right, but only in a static network, i.e., one that is not growing."