The first four essays consider general issues in Internet governance. Gillet and Kapor explain how 99% of the Internet operates without centralised control and suggest that coordination of domain name assignment and IP address allocation will eventually look "a lot more like coordination of the rest of the Internet". Gould provides a UK perspective, suggesting the evolution of English constitutional law and the supranational powers of the European Union as possible models for the Internet. Johnson and Post argue for decentralised governance, highlighting serious problems with alternatives such as extending national sovereignty, international treaties, or governance by international organisations. Rutkowski is less enthusiastic, claiming that "the notion of Internet self-governance is in fact an illusion arising from the relative unfamiliarity of many people associated with the Internet with either its history or the surrounding legal and regulatory constructs in which it exists".
The domain name system is probably the most visible area of contention at the moment; it is the subject of seven of the papers in Coordinating the Internet. Five of them take different positions on the legal arguments over its interaction with trademark law and the political arguments over which organisations should be involved in its administration: the United Nations, the United States Federal government, InterNIC, and the Internet Assigned Numbers Authority just some of the contenders. Taking a step back, Andeen and King compare the DNS with telephony addressing in World Zone 1 (the North American Numbering Plan) and Mitchell, Bradner and Claffy argue that the obsession with the DNS is a red herring, that issues of Internet governance and sustainability are far broader.
Moving on to less prominent but probably more fundamental issues, Rekhter, Resnick and Bellovin write about the need to provide financial incentives for route aggregation and efficient use of the address space, and suggest a framework for property rights and contracts that will achieve this. And Hoffman and Claffy evaluate the advantages and disadvantages of metro-based addressing (based on geography rather than provider).
There are five papers on interconnection and settlement systems and agreements between service providers. Bailey and McKnight survey the variety of interconnection agreements and argue for the introduction of usage-sensitive pricing in conjunction with integrated services with guaranteed quality of service. Cawley, in contrast, extols "the virtues (and appropriateness) of capacity-based pricing and flat-rate tariffs", but suggests that many kinds of pricing models can coexist and that "interconnection and settlement are best left to market forces". Chinoy and Salo review what the transition from the NSFnet to commercial backbones teaches us about oversight and scaling issues. Farnon and Huddle use theoretical efficiency modelling to argue that the present "sender keeps all" settlement system is flawed. And Mueller, Hui and Cheng recount the history of the Hong Kong Internet Exchange.
The final two papers deal with quality of service evaluation, with performance monitoring and statistics. Almes presents an overview of the IETF IP Provider Metrics effort, while Monk and Claffy suggest ways of overcoming practical obstacles — political and economic as well as technological — to adequate data acquisition.
Coordinating the Internet is moderately technical (though it avoids plumbing the complexities of protocols such as BGP) and on some issues its presentation of so many different viewpoints may be more confusing than clarifying. But it is an accessible and informative collection — and an important and timely one, given the increasing urgency of some of the issues it debates.