Milton Mueller <[EMAIL PROTECTED]> wrote:
>That's one of the problems: com, net, and org are *not* structured
>TLDs. NSI stopped any effort to make them into that some time ago. I
>could register in any one of them tomorrow if I wanted to spend
>$70.
NSI's registration policies have set precedents that will be very
difficult to change, even in the fact of alternatives that are
both viable and easy to support within today's DNS architecture.
To take a famous example, registering names like "pokey.org" create a
situation that's difficult to resolve. Names like "pokey" are
popular, and lots of people would like to have them, regardless of
their intent. That creates (financial) value in those names, for those
entities that are willing to pay to get them. In the case of the TM
interests, it makes them feel they must seek legal measures to procure
those names.
With the proliferation of such names, and the relative ease at which
they can be acquired, there is much less incentive to seek out
alternatives, even if they are cheaper (even free!). Having these
names has become "en vogue."
I don't want to go so far as to say that a domain name must map to an
IP address. That would rule out domain names that map to MX records
that forward mail to hosts that are not attached to the Internet
proper, for example. But I do think that the assignment of domain
names has to be made (at least until there are more TLDs) more
carefully than it's been.
A fundamental principle of DNS has been overlooked, imho -- that
it allows for local adminstration of names within established
hierarchies where it makes sense to do so. This was recognized by the
original implementors very early out. A registration of
"www.pokey.vip2.fast.net" could have been handled totally within the
infrastructure of fast.net, and that name is just as visible
throughout the Internet as "www.pokey.org" is.
But as Kent has noted, the genie has been let out of the bottle, and
it won't be easily put back in. I think this is one of the
unfortunate mistakes that has plagued this process.
--gregbo
