Net Nard / The Parts That Must Be Coordinated

Names Have To Be Unique, So Somebody Must Coordinate

The requirement that a name means the same thing everywhere is what forces a genuinely central function into an otherwise decentralised system.

Most of the internet works without anybody being in charge. Naming does not, and the reason is simple. A name is only useful if it takes everybody who uses it to the same place. If two parties could each claim the same name, the name would stop identifying anything, and every user would face the question of whose version they had reached. Uniqueness is not a nicety here, it is the entire point of having names. And uniqueness across a global system cannot be achieved by every participant deciding independently. Something has to keep the list.

So a coordinating function exists. Bodies allocate the top level portions of the naming space, organisations operate registries beneath them, and businesses sell registrations to the public. The structure is layered rather than monolithic, which means the coordination at the very top can stay narrow: it decides what the top level portions are and who runs them, and leaves the running to others. That layering is deliberate. It keeps the irreducibly central function as small as it can be while still delivering the uniqueness the system requires.

The arrangement carries obvious weight. Because names matter commercially and culturally, decisions about them attract argument, and the bodies making those decisions face pressure from businesses, from governments and from users with strong and conflicting views. Those bodies generally operate through consultative processes involving many parties rather than by unilateral decision, which makes them slow and sometimes frustrating. Whether that slowness is a healthy check on a sensitive function or an obstacle to sensible administration is a question on which reasonable people land in different places.