Web names should be divorced from numbers

Technical and marketing issues on the web would be much safer if they were separated, writes Glyn Moody

Technical and marketing issues on the web would be much safer if they were separated, writes Glyn Moody

Previously, I wrote about the unsatisfactory nature of the new domain names proposed by the Internet Corporation for Assigned Names and Numbers (Icann), and the threat of Balkanisation the whole process has provoked. This is not the first time I've criticised Icann and is unlikely to be the last. But I am not alone in saying that the current form of domain name governance is flawed.

There is an entire site devoted to this at www.icannwatch.org/. It is not the even the ravings of someone who has been wronged by Icann it was set up by three respected academics, including Michael Froomkin, one of IcannÕs most intelligent and articulate critics.

There is some necessary background information at www.icannwatch.org/stake/ on why monitoring and reforming Icann is important, and fascinating archives of previous reports and comment on the organisation's latest moves at www.icannwatch.org/archives.

Another organisation concerned by developments is People For Internet Responsibility Opinions at www.pfir.org/statements/policies. But even more interesting are the views of one individual, named Bob Frankston.

Frankston is a father of modern computing. He seems to have been the person who wrote code for the first spreadsheet Visicalc, and has a Web site with essays not dissimilar to those of his former collaborator Dan Bricklin, whose own pages were recently discussed in Siteseeing.

Frankston has been thinking hard about domain name problems. Alongside a good introduction to the Internet Protocol that underlies all packet transport (at www.frankston.com/public/Essays/IP%20An%20Introduction.asp, he has written two essays, Much fuss about the DNS (and IP) and DNS: A safe haven.

His fundamental insight is that the obsession with finding "good" Internet names has led to two different issues being confounded.

The first is technical and has to do with providing content with identifiers, so it can be found. The second is commercial, and is about marketing issues of online visibility.

His solution to the DNS mess is decoupling the two. He recommends the use of "meaningless" identifiers for information like numbers. These would be similar to today's telephone numbers.

Nobody frets much about receiving "meaningless" numbers. To ensure customers use these numbers easily, a variety of directories like Yellow Pages provide a way of contacting a business.

Frankston suggests creating a number of similar online directories, allowing end-users to locate companies. In many ways, these would be like the search engines that are already central to the way people use the Internet.

With the two elements of domain names technical and marketing decoupled, there would be no frenzy to secure the "best" domain names, because nobody would need to see them. Users would turn to directories, particularly if they were built in to browsers. With no demand for particular names, and no demand for new hierarchies (which would be irrelevant), the names "crisis" would be shown for what it really is the creation of those with vested interests in building flourishing businesses based on artificial scarcity.

Frankston's analysis also explains why Icann is such a shambles. It has pandered to various commercial forces brought to bear on it, and inevitably comes up with a compromise pleasing nobody.

The time has come to give responsibility for running this fundamental aspect of the Internet back to the people who created the it Ð the geeks. We must allow them to use pre-existing neutral structures such as the IETF that have served it so well. That way, solutions will be chosen on what is best for everybody, not just greedy dotcom entrepreneurs.

Read more on Managing IT and business issues