[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Comment-Ip] WIPO Report is 100% backwards and should be rejected entirely

This is a comment on the "Final Report of the WIPO Internet Domain
Name Process April 30, 1999".  I believe the report should be
rejected in its entirety, for reasons specified below.

WIPO proposes a significant expansion in intellectual property rights,
at the expense of free speech and due process of law.  The entire WIPO
proposal should be rejected, and if a new proposal is needed, it
should be drafted by a neutral third party (not one with a vested
interest in the expansion of intellectual property rights).

The entire proposal suffers from a congenital confusion between
trademarks and domain names.  DOMAIN NAMES ARE NOT TRADEMARKS AND DO
completely settled law.  (The WIPO Proposal is an effort to change
that law, de facto if not de jure.)  It is possible for a domain name
*holder* to infringe a trademark, just as it is possible for a
newspaper publisher to infringe a trademark.  But a domain name
in-and-of itself cannot infringe, just as mere use of a phrase such as
"Star Wars" cannot infringe.  The existence of occasional *people* who
infringe trademarks does not justify closely regulating newspaper
publishing, or domain registration, on the grounds that trademarks may
be violated.

Domain names are, like other words and names, intimately tied to
freedom of expression.  The *name* of a publication is just as
protected by freedom of expression as the rest of its content.  We
have seen a long history of deliberate use of intellectual property
law to bludgeon ordinary speakers into giving up their right to speak
on the Internet in the manner in which they choose -- even when there
is no conceivable trademark infringement occurring.  For example, as
briefly cited in paragraph 324, a toy company persisted in
administrative and legal proceedings to shut down "pokey.com", a site
run by a small boy nicknamed "pokey" from birth, until a last-minute
appeal to the famous animator who originally created the trademarked
"Pokey" character finally resulted in the company reining in its
abusive lawyers and showing common sense.  Thousands of similar cases
have occurred which did not make the newspapers.  The WIPO report
conveniently plays down its few references to the widespread abusive
practices on the part of trademark holders who wish to expand their
control over language far beyond the realm of "preventing consumer
confusion regarding commercial goods", the proper realm of trademark
law.  (Control over language is usually called "censorship", but the
WIPO report is careful to avoid mentioning that word.)

WIPO proposes that domain names be denied to those who wish not to make
themselves available for nuisance suits by IP fanatics.  They state that
users who register a name and pay the required fees must be denied their
chosen domains for lack of personal contact information.  Trademark
holders who seek to enforce their rights must bear the cost of enforcing
their rights, including the cost of finding and serving the parties
involved.  The cost of finding these parties should not be "shared"
by requiring every possible party to such a lawsuit to publish contact
information enabling them to be conveniently sued.

Further, the WIPO proposal states that rather than permitting a court
to look at the issue, the entire matter should be handled by an
administrative process managed by WIPO itself.  The validity of this
administrative process would be suspect, therefore the proposal would
*require* every domain name users in every jurisdiction to enter into
a contract prejudicial to their own rights, explicitly permitting the
process to be used against them.  This process need not require any
malfeasance on the part of the end-user who registers a domain name;
all it requires is "an allegation of abusive registration of domain
names that violate trademark rights".

The proposed definition of abuse includes an element that if the
end-user who registered the domain has offered to sell it to a party
with a trademark, abuse has occurred by definition.  Suppose that I,
the end-user who registered "toad.com" long before Internet Service
Providers existed, were contacted by a company which uses "Toad" in
its trademark, declined to sell the domain to them, and then later
changed my mind and offered them the name.  At that point, under the
proposal, I have become abusive, and therefore can have my name taken
away at the whim of any "toad" trademark holder.  (The essential
elements are that there is a similar trademark (there is); that I have
no trademark (I don't), and that I have offered to sell it (I have).)

Trademark law already recognizes that many, many parties are free to
use the same mark concurrently.  Even within a single national
jurisdiction, scores of parties can share trademarks in different
fields of business; and there are hundreds of national jurisdictions.
Nevertheless, the WIPO proposal continues to refer to "the" trade or
service mark (e.g. 171(2)(a)), in a misguided belief that a single
powerful trademark holder can compel all others to forego the use
of a word.  This unconscious emphasis on a completely flawed model
of trademark law pervades the document.

Just as shared use of marks is the proper legal result in today's
ordinary trademark law, shared use of marks is the proper result in
domain names.

The entire phenomenon of cyber-squatting exists because of the
artificial historical limitation of the DNS to one large domain (.COM)
and two ancillary domains (.NET and .ORG).  Similar predatory
practices are not prevalent in other naming realms, such as in US
states' corporate name chartering or in trademark registration.
"Squatters" do not register likely future corporate names in various
states and then attempt to blackmail new companies; nor do they
commonly register trademarks that they anticipate others will want
later.  This is because there are many variations of the desired
trademarks and corporate names available; the intended "victim" of
such a squatter need only change a few letters, or add a word to their

ONE, OR SEVERAL, SECOND LEVEL DOMAINS.  If hundreds of TLDs existed, a
cyber-squatter would either have to register unprofitable names in
hundreds of TLDs in order to sell a name in one; or would have to
guess which TLD a future trademark holder might want to put its mark

The issuance of domain names, just like trademarks and company names,
should be done on a first-come, first-served basis.  Period.  Without
any administrative way for some third party to come along and cancel
them.  The only way for a third party to compel a change in this
contractual relationship should be to have a court issue an order,
after weighing all factors and providing for an adversarial process
to encourage the truth to come to light.

A company need not register its marks in every possible top-level
domain in order to protect them (and very few companies do today).
Any third party's domain name which actually infringes the first
party's trademarks by creating consumer confusion can be dealt with
under existing law.  Domain names in TLDs not used by the first party,
registered by other parties, which do not infringe the first party's
trademark, by definition do not create confusion with a mark protected
by trademark law (or else they'd infringe).  Therefore they cannot
legally or morally be interfered with by the trademark holder.
Creating an administrative system that would permit trademark holders
to effectively interfere with such other non-infringing parties
offends not only free expression, but trademark law itself.

Similarly, WIPO proposes a large expansion in the control provided to
owners of "famous marks".  Special protection against the "dilution"
of famous marks is a new feature of trademark law, and only applies in
a small number of jurisdictions.  (The treaties cited by the WIPO
paper, with hundreds of signatories, do not result in blocking
noncommercial names, or names which do not cause actual confusion with
goods and services being offered in commerce by the "famous" mark's
owner.  US "dilution" law is much broader than these treaties provide
for.)  Early experience with this law in the United States reveals
that virtually *every* company involved in trademark litigation claims
to have a "famous mark" because it has little or nothing to lose by
making such a claim, and much to gain.  A right envisioned as
protecting a few dozen or perhaps a hundred truly famous marks has in
actual practice been claimed by every Tom, Dick, and Harry.  WIPO
continues this tradition by placing no limits on the set of marks
which could be protected by this administrative scheme, in fact
proposing that it not even require that the "famous mark" be

WIPO's response to this concern is in paragraph 270, where they state
that some commenters have this concern, and then "rebut" it by merely
stating that they don't agree.  Besides being insufficient, it is
wrong.  There is no need to "give expression to" trademark law in the
DNS.  Courts have already "given expression to" trademark law in the
DNS, and in every other realm in which trademarks can be diluted or
infringed.  What the major trademark holders behind WIPO attempt to
administratively "give" themselves in these proposals is far beyond
what trademark law and the courts have already granted them.

Pargraph 285 is particularly notorious.  It defines the circumstances
in which a mark is considered "well-known" and thus eligible to censor
other users, by circular reference: If anyone else wants to use the
mark, then it must be well-known, therefore the "anyone else" parties
can be rebuffed.  Note again the one-owner-per-mark mentality in the
mention of "non-authorized" parties, making the assumption that PRIOR
to the determination that a mark is famous, a single party will
"authorize" uses of the mark.  This directly flies in the face of
existing trademark law and practice worldwide.

WIPO's proposed "evidentiary presumption" sets the stage for a further
expansion of the rights of large companies, to the detriment of
ordinary speakers and small businesses.  The presumption only applies
in administrative proceedings, and there is only one kind of
administrative proceeding specified -- cancellation for "abuse".
(Though somehow well-known marks are censored by some sort of
administrative panel, this isn't a "proceeding"?)  In an actual case
of abuse, there is no need for an evidentiary presumption.  The entire
point of abuse cases -- as proposed -- is that they are 100% clear-cut
according to straightforward rules, and do not require the weighing of
any evidence.  Therefore, why does this provision exist?  This
provision is a Trojan Horse that will open later, once trademark
interests have expanded the realm of administrative procedures, as
contemplated in e.g. paragraph 168: "We are persuaded by the wisdom of
proceeding firmly but cautiously...After experience has been gained
with the operation of the administrative procedure...the question of
extending the notion of abusive registration to other intellectual
property rights can always be re-visited."  It should be rejected.

Regarding the expansion of the set of TLD's.  Paragraph 310 is wholly
without merit.  I'll just point out some of the best parts.

(i) is technically challenged.  I am intimitely familiar with the
technical operation of the Domain Name System, having authored some of
the software in BIND, and having set up and run a reliable set of
worldwide SLD servers for CORE.  There is no technical problem with
having hundreds, thousands, or ten thousand top level domains.  Today
we have a single second-level domain that contains several million
entries, served in fact by largely the same set of servers that also
provide the root domain -- and the domain name system works fine.
This is a complete bugaboo, exacerbated by the amazing non-sequiter in
the last sentence, which somehow confuses the root servers which serve
online name-lookups with the Shared Registry System that arbitrates
new registration requests within a single TLD.

(ii) is also specious; it claims that we now have "competition"
in domain registration because five entities are engaged in a trial
in which they each send money to Network Solutions.

(iii) somehow suggests by hundreds of words of circuitous language
that some words mean more to humans than other words.

(iv) refers to the useless suggestion that ".sex" be somehow made a
ghetto which sexually related materials are forced into and which
minors are somehow prohibited from entering.  Neither half of this
proposal is remotely legal or constitutional in many jurisdictions,
nor would either half be a good idea.  It's not clear why WIPO refers
to it, except as a strawman for the uninitiated.

Paragraph 310 completely ignores the free expression "way of viewing
the question" as it enumerates ways of viewing the question of having
more TLDs.

Paragraphs 315-320 do not relate at all to the question of whether
additional TLDs should be created; instead it focuses on cases
where people have registered misspelled or similar names in .COM.
It's not clear what these are doing in this section at all, other
than to provide someplace for trademark holders to gripe about
domain names.

Paragraph 321 quotes someone from a company that "I am spending more
money on Internet-related issues than I am on consumer products piracy
in the United States, and that really is unbalanced in my view."  What
this has to do with other TLDs is unspecified.  This company's
spending is unbalanced in my view as well -- they should stop trying
to "keep the domain name register clear of names registered by others
incorporating our famous trademarks".  No wonder it gets expensive,
when you are trying to enforce rights in the courts which were not
granted by the enabling legislation.  Trademarks do not grant a
company the right to prevent all other speakers from using the marks;
only to prevent confusion with a product or service.  By the way, this
company is of course another Tom or Dick with "famous trademarks" that
deserve special protection.  In their opinion.

Paragraph 323 slantedly refers to trademark owners as "victims of
domain name abuse", while ordinary citizens who lost their legitimate
domain names due to abuses by trademark holders were merely affected
by companies who "resorted to defensive practices that those
commentators find undesirable".  I find the entire WIPO proposal
not only undesirable but also abusive.

In summary, this entire section provides no particularly strong
rationale for limiting the set of new TLDs.  It makes no case
that intellectual property protection would be harder with more

In fact, the lack of a large number of TLDs has created most of the
problems which the WIPO Report seeks to address.  But the WIPO authors
are too blinded by the large-company ideal of one-mark-one-owner to
be able to see the real benefits provided by an open market in 
domain names that would permit hundreds or thousands to share the
same second-level mark without confusion.

Paragrpah 348-350 refers to "keyword" systems such as Netscape's.  In
fact I believe that Netscape implemented this system in an attempt to
obtain some of Network Solutions' revenue windfall for itself.  As a
CORE member, I proposed to Netscape that they should charge NSI
millions of dollars per year to keep the feature in their browser that
looks up unknown domain names in NSI's proprietary ".COM" domain.
Instead, perhaps after a fruitless sales call on NSI, Netscape decided
to implement *their own* database of names that companies would have
to pay to register in so that browser users would be able to easily
reach the companies.  No wonder it has the same perceived problems
as the domain name system -- it attempts to reproduce it, just
for another party's benefit.


	John Gilmore
	PO Box 170608
	San Francisco, California, USA  94117
	+1 415 221 6524	voice
	+1 415 221 7251	fax