I have somewhat of an alternate take on Stef's approach, which I'll
try to explain below.
A first principle: Fully qualified domain names are sequences of
octets (components) separated by special qualifiers (dots). We
call the rightmost component the TLD, the component to the left of
that the SLD, etc.
There exist a finite (but large) set of domain name components. Thus,
the set of components is enumerable. It could in theory be listed in
any zone.
I offer that what a registry is doing is publishing a (virtual) set of
components that are available to be used as references to network
resources. Registrants contract with the registry to establish
mappings from these components to resources that they want to
associate with the concatenated set of components that comprise that
particular component and anything higher in the tree (in other words,
the fully qualified domain name).
The People's argument: I see that there is a component in your table
that's not claimed. I like it and I am willing to pay for it, so I
contract with the registry to establish the mapping I want. The
contract may be terminated by either party at any time.
The Cybersquatter's argument: Same as the People's argument, with one
extra claim: If I choose, I may sell this contract to a third party.
The third party has no right to demand that I voluntarily give my
contract up.
The TM holder's argument: Some of the components you have listed are
names that I have some vested interest in. Therefore I am the only
one with any legitimate claim to use those for mappings to network
resources. No matter who established contracts earlier, once I decide
I want that component I should be able to get it.
My general feeling is that a situation has arisen where it is
considered attractive to have names with three components or less.
A highly sought-after component sequence is www.component.tld.
This has placed intrinsic value on the set of components to the left
of the rightmost dot. In addition, since some of those components
already represent strings that have some intrinsic value derived from
outside the context of domain names, the people who have interests on
those names feel they are justified to get those components.
So here's some questions to ponder. Are SLD's "special" in that they
should be "reserved" for certain types of uses? Why is that? Let's
say that some 3LD gets very popular, like
famouscompany.sub-registry.tld-registry. Does FamousCompany have
rights to that FQDN? I think this is a key point that was brought up
earlier (I forgot by whom).
I also think that Kerry's comments about the TM interests deriving
value from character strings has some merit. What the TM interests
are doing is trying to establish some claims over particular character
strings. The value they derive is their ability to have their
subscribers refer to their network resources by easily remembered
names, such as www.mycompany.tld.
I also think that if we are talking business models here, consider
that this is just one business model that happens to have a certain
amount of support, but it is not the only one, and not necessarily the
best one. The model of web hosting sites like GeoCities might evolve
as a more effective model of establishing online presence. Or a newer
model might emerge, based on a new technology, like "We're in the
RealNames Pages under the United Van Lines index." So it might not be
wise for ICANN to invest in any business model based on TLDs, because
the marketplace might choose another.
--gregbo
