Chris Ambler in his blog entry "Standard Roots?" responds to my question why should DNS depend on a single source of authority while the internet's time protocol, NTP, survives quite well without any central authority. In particular he asks how one ought to determine which is the more rightful owner of a domain, such as my .ewe TLD if there are multiple claimants.
My answer is very simple: "The Old Fashioned Way". There is a well traveled road formed by existing practices of competition assisted by the traditional laws governing trade and service marks.
To be more precise: Imagine that two .ewe's are started by different people. There would be the normal race to establish territories, to obtain trade and service mark rights (common law and also by registration), and, most importantly, to obtain customers. Within the spheres established by the scope of their trademark rights they will evolve over time (which will also stretch and shape the scope of the marks). And over time they will each exercise greater or lesser degrees of creativity or define more or less interesting products and prices. The economic powers of each will be shaped by their success within their respective marketplaces. If they come to blows over trade or service mark rights then they can fight it out in the courts or one can acquire the other.
The point is that there need not be any overlord of names that says that my .ewe wins and that your's looses. Instead the normal rules of economic competition and trademark law determine the outcome.
You might ask a few questions: What if the names are not used in commerce and thus not subject to the law of trade/service marks? What about the notion of first come first served? What about user confusion?
As for the first question: If two combatants for .ewe want to duke it out without resorting to trademark law then I believe that we may want to fall back upon the old rule that first-in-time is first-in-right. But the important point here is that the choice to apply that rule is not made by some singular authority of names but rather directly by each person who constructs a root zone file and indirectly by each person who chooses which root zone (and thus root zone file) to use when he/she constructs his/her view of the internet landscape.
By analogy: there are millions of speed dial buttons on millions of telephones around the world. The telephone systems of the world do not wobble into oblivion because each person gets to chose what telephone numbers are assigned to the buttons on their own phones. We neither need nor want a worldwide central authority over speed buttons - but that is what the existing DNS policy regime is tending towards.
As for the second question: First come first served is a principle of trademark law already, so if two marks associated with same-named TLDs should come into conflict then temporal priority is a factor. If the names are outside of the trademark space then those who build root zone files will probably often take time precedence into account among the factors they evaluate when choosing the sources of the TLDs they wish to incorporate into their root zone's portfolio of names.
As for the last question: It has often been asserted that users of the internet will be confused if domain names have different meanings in different places. I don't find that to be a particularly valid argument when I reflect upon the fact that domain names already change meaning over time, often over very short periods of time. Domain names do not have temporal stability.
In addition, users might actually find the net to be more "stable" from their personal points of views if they can shape the landscape of names by saying what names they chose to recognize and what names they chose to exclude. Today's DNS policy regime is based on a Procrustean view that imposes a catholic name space upon everyone whether they want it or not. That kind of dictatorial approach is not unlike a policy that would require everyone on the internet to speak using only the English language and character script.
The bottom line of all of this is this: We don't need any new governance body to resolve (pun intended) between two .ewe's. This kind of problem has both existed and been solved for ages. The marketplace and legal mechanisms to reconcile these disputes exist and are known to work without the need for new bodies of governance.
Posted by karl at February 23, 2005 6:31 PM