Re: [wg-c] A branded TLD would be .nsi

[Jonathan Weinberg <weinberg@mail.msen.com>]

	I've not gotten involved in this debate b/c I think the point is
tangential, but . . .

	As Kevin pointed out a while ago, there are serious obstacles in the way
of a TLD becoming a service mark for domain name registration services (or
registry maintenance services).  Here's why: A trademark (or a service
mark), under U.S. law, is a source identifier.  That's what it's for.  If a
string isn't operating as a source identifier, then it's not operating as a
trademark.  Generic words (like "chair," when applied to chairs) can't be
trademarks because they can't be source identifiers; the ordinary person
understands them to refer to the whole class of goods, not to goods
emanating from a particular source.  (Of course, "chair" can be a perfectly
good trademark when applied to gumdrops, say, but that's another story.)

	Let's say I start the TLD .www.  Can I claim that string as my service
mark?  The problem, it seems to me, is that the string is a generic
descriptor of "registry services in the .www TLD."  Because only one entity
can provide that service, the consumer doesn't need a source identifier.
Nor does the string work as a service mark for registrar services.  If
there is only one registrar, we have the same problem as above, and if
there are many, then the string doesn't identify any one of them in
particular.

	Now, it's true if consumers believe that I am the provider of
registry/registrar services for .www, but ICANN gave the .www franchise to
somebody else, ordinary consumers might be confused by seeing the .www
string out there.  I could bring a trademark lawsuit, under U.S. law,
seeking an order that would end that confusion.  The real question is
whether the court would respond by forbidding the new folks to offer
services connected with the TLD at all, or whether it would figure that it
could avoid any confusion simply by ordering them to prominently display
the words "NOT AFFILIATED WITH JON WEINBERG" on their web site.  *That* one
is highly dependent on the facts.  We get the same problem if Joe Blow
starts the .ibm TLD; people might be confused, thinking that Joe's registry
services were somehow emanating from, or sponsored by, IBM.  IBM, again, is
entitled to an appropriate remedy to avoid confusion.

Jon


Jon Weinberg
weinberg@msen.com


At 08:21 PM 8/19/99 -0400, Martin B. Schwimmer wrote:
>I'm not sure a point by point discussion is productive but when discussing
>the "branding" of TLDs let's keep clear several distinct questions:
>
>1.  Whether .com is today a trademark of NSI;
>
>2.  Whether .web is today a trademark of CORE or IOD or no one;
>
>3.  Whether TLD suffixes can theoretically function as trademarks;
>
>4.  Whether TLD suffixes should as a matter of DNS policy be allowed to
>function as trademarks.
>
>The easiest question is no. 3, because it is certainly possible to posit a
>set of facts where a TLD suffix functions as a TM (the easiest being to
>take an existing TM - .mci, the other being taking an arbitrary term, .xyz,
>and ICANN or whoever decrees that only X can be the register/registrar of
>.xyz).
>
>Questions 1 and 2 are mixed questions of law and fact I leave for the courts.
>
>Question 4 is not really a TM question but a DNS policy question and I
>refrain from putting in an opinion.
>
>As to the other points, I did type in "identifier" when I meant to say
>"administrator" so my point, right now, with these gTLD suffixes, they do
>not by definition identify an adminstrator, they, by definition, identify a
>domain.  As to arguments as to whether NSI has achieved secondary meaning
>in .com, I think there is a pelthora of arguments to the contrary - the
>advent of multiple registrars in .com pretty much being the show-stopper.
>
>As to whether Sally Abel ever argued that a TLD cannot be branded as a
>matter of metaphytsical trademark law, I believe that she was recently
>quoted in either wired.com or news.com on the CORE/IOD matter, and that she
>implied that there were plausibly circumstances where a TLD suffix could
>function as a TM (and she was careful not to take sides).  Again, this is
>different from a non-TM position as to the desirability of branded TLDs.
>
>
>
>
>
>
>
>
>
>
>At 06:20 PM 8/19/99 -0400, you wrote:
>>
>>
>>Martin B. Schwimmer wrote:
>>
>>> MM wrote:
>>
>>> >Both Ambler and NSI want to be the exclusive registry
>>> >for a specific TLD string, .web and .com respectively.
>>> >Those claims as I understand them in no way involve
>>> >a claim that the string ".web" or ".com" belongs to them,
>>> >whenever they are used. It is simply a claim that
>>> >no one else can register names under .web and .com on
>>> >the common root of the Internet. It is a claim to exclusivity
>>> >of a certain type of service as branded by the TLD.
>>>
>>> As branded by the TLD administrator's trademark.
>>
>>Not necessarily. Substitute the word "identified" or"distinguished" for
>the word
>>"branded" above and my point is
>>clearer. The .web registry is basically the *only* place
>>that knows how to resolve a domain name ending in .web.
>>Therefore any domain name ending in .web is associated
>>with the services offered by that registry. IOD or someone
>>else can obtain trademark rights in .web by operating
>>this registry. It is an identifer that distinguishes his business
>>from another. They could argue on TM grounds, it seems
>>to me, for exclusivity of registration services under .web--
>>but not, of course, for exclusivity of the use of the term
>>web or dot web.
>>
>>
>>> Ambler's and NSI's claims
>>> would rise or fall on issues other than trademark law.
>>
>>Based on the rationale above, why?And if you're right, which issues?
>>
>>> Not analogous, as in this case the DN owner doesn't wear the suffix,
>>
>>Aren't we being a bit too literal, here? It "wears" the suffix onits full
>domain
>>name.
>>
>>> often incorporates the TLD suffix as part of its mark, as in amazon.com or
>>> drugstore.com or 1-800-flowers.com. PEPSICO, AMOCO, CITICORP and my
>>> favorite, ESPRIT DE CORP are all trademarks which incorporate corporate
>>> suffixes, which suggests why the suffixes themselves couldn't function as
>>> identifiers of a single source.
>>
>>But a .com suffix clearly *does* identify a single source for
>>registry services
>>
>>> >I would still like to see a TM lawyer explain why
>>> >TLDs can't be branded.
>>>
>>> Has a TM lawyer argued that they can't (as opposed to arguing that they
>>> shouldn't, on non-TM grounds)?
>>
>>Yes. I tihnk Sally Abel did. More significantly, I would have thoughtthe
>TM/IP
>>constituency would be more sympathetic to the notion.
>>
>>> No.  TLD, by definition, stands for top level domain.  .edu identifies the
>>> domain, not the domain identifier.
>>
>>I don't understand this distinction.
>>
>>>  The adminstrators of .com, .edu, .org
>>> and .net are not the same in 1999 as they were in 1991, and yet the TLDs
>>> themselves are unchanged.
>>
>>The owners of brands may change? TMs can be bought and sold?
>>
>>> I didn't say that TLDs couldn't theoretically be branded.  They could be
>>> branded. I understand that you are a big advocate of branded TLDs
>>
>>I am a big advocate of a non-uniform, heterogeneous modelfor DNS
>administration.
>>Let there be shared generics, branded
>>ones, and everything in between.
>>
>>
>>
>>
>
>@ @ @ @ @ @ @ @ @
>
>
>