Re: PAB John Jacob Jinkleheimer Schmidt.

John C Klensin (
Thu, 05 Mar 1998 07:51:51 -0500 (Eastern Standard Time)

On Thu, 05 Mar 1998 10:40:12 +0000 (GMT) Jim Dixon <>

> If I understand you correctly, you are saying that people who
> register names at the second level under .nom should be required
> to accept registrations at the third level, and so forth. This
> would be a misreading of RFC 1591 and a legal can of worms.


Having been slightly involved in the writing of the relevant
text in RFC 1591, I don't believe this would be a misreading of
the language or intent of that RFC as far as it goes. Of
course, 1591 was written for a very different purpose, and
anticipated that TLDs would be either "generic" and categorical
(i.e., not "open" except to members of the category) or
country-based and with some coherent structure intra-country.

"smith" is not a category within the mathematical definition of
that term, it is just a name shared by a lot of substantially-
unrelated people/entities as well as a number of clusters of
related ones.

I cannot disagree about the cans of worms associated with
trying to enforce strict hierarchy on the first person/group to
obtain [most] human surnames. The legal one is just one of many
such worm-collections.

> Experience at existing registries has shown that trying to enforce
> policies about names registered has led to endless hassles. The
> best policy is to ignore the actual strings registered, insofar as
> trademark and IPR concerns make this possible.

Personal opinion: SRI used to promote and encourage the rules,
although not always consistently to put it mildly. NSI started
out a little sloppy about those efforts and then apparently
gave up entirely when the volume started to rise (for all I
know, that timing might have been coincidental -- I'm not going
to get into an argument about causes here). Giving up set, IMO,
unfortunate precedents.

While encouraging cooperation and goodwill worked better in the
Internet's pre-industrial period than today, I believe that in
circumstances where the distinctions are plausible, clear
statements of intent and some gentle educational effort when
there is evidence that people don't understand the intent will
still cause the right thing to happen most of the time. I'm not
suggesting legal-level controls or ones that can successfully
resist determined attacks by sociopaths or deliberate wreckers,
only that clear intent and consistent requests for cooperation
may yield 75% solutions, especially when there aren't other
constraints (e.g., readings of the trademark laws that require
strong defenses against "dilution"). And I think that, in
these sorts of cases, those 75% solutions may be worth the