Thanks for the thoughtful, well articulated analysis.
Thinking back to the earliest days of ARPANET development, when surely the
property rights of which you speak must have sprung, can you find any record
of any awareness of 'ownership' of the identifiers used in the network? Did
the operators of new nodes have to pay somebody for their identifying
number(s)? Even though an 8-bit address field limited the number of sites
The concept of names for machines seems traceable to Peggy Karp's
RFC in 1971, although Bob Kahn in a recent conversation suggested
it goes back further to informal lists kept at the SRI NIC. Peggy's
RFC in any case established the first formal use of names and the
host table as a lookup mechanism. Typically, as now, one just informed
an entity of the desired name, which was then made available on a
machine as part of a table. There is no significant difference today
except the maintenance has largely been pushed out to the network
edges with local zone files.
I don't like ICANN all that much either, but consider this a plea to those
on all sides of this debate to avoid using concepts of ownership in their
analyses and refer to concepts of use (yes, even exclusive use), instead.
What we're all trying to figure out is who has (or should have) the
authority to make exclusive assignments of Internet identifiers (and on what
terms), not who owns what. The difference is fundamental to the nature of
the Internet (think Open Source vs. proprietary code) and will play out in
many aspects of Internet architecture and policy as the network continues to
develop.
I would argue that no one should have "the authority to make exclusive
assignment of Internet identifiers." Indeed, there is no such thing.
You can today use any identifier you choose - and many institutions do.
However, unless you have made special arrangements, your traffic might
not end up in the right place. As a shared user network, the users
vote as to whose identifier system is used and on what terms, not
some higher authority - ICANN or otherwise.
--tony
