Fast movers are slow to keep up

Back in the days of the dot-com boom, people talked of "internet time" to express the notion that governments and regulators could not possibly hope to keep up with developments on the net. This book is an excellent demonstration of the principle.

On October 28, 1998, President Bill Clinton signed the Next Generation Internet Research Act, which mandated a wide-ranging study into the operation of the domain name system (DNS) - the mechanism that performs the translation from the user-friendly www.example. com names into the internet protocol (IP) addresses that are used to contact websites. Eventually, a committee of the internet equivalent of the great and the good was put together by the US National Research Council. This held seven meetings in 2001-02; a report was eventually written in 2004, extensively reviewed by practitioners in early 2005 and finally published as this book last September.

Out in the real world, there had been two presidential elections, four congressional elections, one millennium and seven Super Bowls. But on the net things were moving even faster and 1998's problems were tackled long ago.

The report starts with an 18-page "executive summary", clearly aimed at executives with time on their hands, which can be summarised into "the DNS is really useful, it ain't broke and ICann should be made independent".

(ICann is the Internet Corporation for Assigned Names and Numbers.) The lack of anything that is really broken makes it hard for the report to rise above mere description. However, where there is controversy, for example, on how the ICann's independence can actually be achieved, the committee have little to say that is more than mundane.

The strongest part of the report is the discussion of the technical aspects of the DNS. It contains one of the best accounts I have seen of the practical failings of its predecessor, a single file called HOSTS.TXT that listed every machine (all 562 of them) on the old Arpanet - from which the internet evolved. It is also good on the mechanics of the current system and the operation of the root zone, which lists the 15 or so global top-level domains (TLDs) such as .com, .net, .org, .biz and .info; along with several hundred country codes such as .uk, .de, .fr and so forth.

There is also a reasonable account of the types of dispute that have arisen with the registration and the use of domain names, especially those involving trademarks, although one should not be surprised that the analysis and legal background is almost entirely US-centric. But the American bias is not universal, and there is a detailed account of the technical developments by which the DNS is being extended to permit countries to have meaningful domain names in their local scripts (Chinese, Korean, Arabic and so on).

In contrast, the material on internet governance is disappointing. Since the start of the internet, ultimate control of naming has been with the US Department of Commerce, but ICann was created with a view to having it take over this role. Initially they did well by arranging for competition among domain name registrars (hence domain names now cost $5 (£2.80) rather than $70), and they set up a functional scheme to stop cybersquatters from registering domains and holding trademark holders to ransom. However, in 2000 they rejected 35 TLD proposals in controversial circumstances.

Later reforms of membership (abolishing elections after just one occurrence) and TLD approvals (we will shortly see .cat for Catalan websites - albeit with feline discussion forbidden unless in that language) have improved ICann's standing a little, but it remains subject to extremely fierce criticism.

The report recommends that ICann continue to be encouraged to adopt complete independence, but although a number of governance models are discussed, the committee avoids controversy and sometimes even an opinion, so there are few firm conclusions on what should be done. However, it is clearly pointed out that without initial agreement as to the tasks ICann should perform, there is little hope of establishing a suitable structure to manage it.

And here, on the topic of ICann, once again the report falls victim to "internet time", because just weeks after its publication came the biggest diplomatic row yet, with the European Union joining developing nations to press for international control of the internet. Yet in a last-minute deal in Tunis, November's World Summit on the Information Society set up an apparently toothless Internet Governance Forum, leaving the US in full control of ICann, and hence the DNS, for the foreseeable future.

The report is currently available as a free PDF download from the US National Academies Press website - which means value for money and timely access. I do not think Congress got the same.

Richard Clayton is a researcher in the Computer Laboratory, Cambridge University.