From: Steve Kent <kent(_at_)bbn(_dot_)com>
To: pem-dev(_at_)tis(_dot_)com
This wire service news story illustrates the principle that
one cannot simultaneously have very short names and grow the system to
accommodate lots of organizations. I expect this is just the first of
That's nonsense. Telephone numbers are very short names and
accomodate essentially all the organizations in the world. You must
mean something else. Like the shorter the names you use and the
larger the number of different names you use then the smaller the
average distance will be between said names in some sort of name
recongition space.
So what?
People can choose how short or long their names are in the DNS. If
they want to, they can register the name as a trademark in many
different ways before and/or after they register it as a DNS name.
In many cases, the name may already be registered as a trademark,
corporate name, individuals name, etc.
There is no problem here.
what will become a torrent of legal conflicts over DNS names. It's
I don't believe there will be such a torrent. If there are a few, so
what? It's not a problem for the Internet or Internic. DNS names are
public. In the USA at least anyone can sue anyone in court for
anything. There is no reason that DNS names can't end up being
treated pretty much like trademarks. If people want to negotiate or
argue or arbirate or go to court over such things, when the resolution
conflict mechanism chosen comes to a decision, the DNS can be
adjusted.
There are lawsuits over trademarks and corporate names all the time.
Verbosity is no guarantee of conflict avoidance. Trademark law has
lots of useful principles in this area. Such as allowing
"conflicting" trademarks to be simultaneously used for businesses of a
"local" nature if they are geographically separated. (For the
internet the distance should be measured in cyberspace rather than
geographcially.) Or the division of trademarks into three classes of
decreasing strength: made up new words like Exxon which are given the
broadest protection, even to some extent against use in unrelated
fields; fanciful marks which are existing words but used in a new
sense, like a Prayer brand computer; and the weakest category,
descriptive word which have gained a secondary meaning, for existing
words used in a normal sense but which have been used so extensively
as to gain a secondary meaning. (For example, if you manufactured a car
and called it "The Automobile" and invested a tremendous amount in
advertising so that people tended to recongize that capitalized phrase
as identifying your product, you might be able to get some very narrow
protection against people who were clearly trying to cash in on your
advertising but you could not stop people from the normal use of the
word "automobile".)
The analogy I am about to give is not very exact but I feel that your
position is the same as that of someone that says only the correct
internationally agreed to chemical nomenclature should be used. Too
bad that it may be thousands of letters long sprinkled with numbers,
hyphens, parenthesis, etc. After all, its the international standard
and unambiguoulsy identifies the chemical.
My position is that no one but a tiny spinkling of scientists will
ever use such names. Manufacturers can choose brand names for their
chemicals, and they do, and there are short generic names. Those
short names are the only names that 99.99%+ of the people.
Trying to force impractical the long names like DNs down people's
throat will only lead to a welter of conflicting short aliases in
different communities.
Actually, DNs are a lot worse than chemical names because the set of
elements actually included is not standardized. Given a complete
dossier on a person or corporation, you really still have no idea what
the DNs they might appear under in the fictious The X.500 Directory.
You have a much better chance given the molecular structure, of
getting the chemical name.
(The whole X.500 "standard" is a lie in that it provides just enough
framework that you can claim there is an single international standard
distributed directory while providing so much freedom to the
administrations within it that with current technology it is not
practical to produce a client that can deal with the resulting mess in
a useful way.)
not that DNS names cannot be structured to provide a reasonable degree
of descriptiveness through the use of a few layers of geopolitical
qualification and use of longer strings. The DNS name
"cnri.reston.va.us" is a step in the right direction, but its unusual
among domestic (U.S.) entries and even it does not spell out the full
name of CNRI. It is, however, the case that the Internet community
has not forced such structuring of DNS names and thus we have problems
So when are you going to force International Business Machines
Corporation (or is that Corp. or Incorporated or Inc. ...) to stop
using IBM? When are you going to force me to always includes my
middle name spelled out in full (which does not help since there are
at least three living persons named Donald Eggleston Eastlake and who
is to say that one might not be referring to someone who is deceased)?
of the sort described here.
Let me say it again very clearly: There is absolutely no problem
here for the Internet.
There is a balancing between the benefits and penalties of very short
names, like dotted notation IP addresses, and grotesqueries like DNs.
DNS names are near the optimal middle ground. DNS and RFC-822 email
addresses based on them are *the* names of choice for good reason. We
should build on their success rather that push either very short
(i.e., serial numbers) or very long names (DNs).
Steve
Donald
==================================================================
Name Flap On Internet Raises Question Of Trademark Law
By ELIZABETH WEISE= Associated Press Writer=
SAN FRANCISCO (AP) _ Brand names and trademarks have long been
so important to companies that they will protect them in court.
Now, it appears, so has a computer address.
...