ICANN ICANN Email List Archives

[irt-draft-report]


<<< Chronological Index >>>    <<< Thread Index >>>

IRT Comments by Leap of Faith Financial Services Inc. (May 6, 2009)

  • To: irtp-draft-report@xxxxxxxxx
  • Subject: IRT Comments by Leap of Faith Financial Services Inc. (May 6, 2009)
  • From: George Kirikos <gkirikos@xxxxxxxxx>
  • Date: Tue, 5 May 2009 23:16:16 -0700 (PDT)

Comments on IRT Draft Report

Submitted By: George Kirikos
Company: Leap of Faith Financial Services Inc.
Company URL: http://www.leap.com/
Date: May 6, 2009

I. Introduction

1. We reiterate all previously submitted comments on the new gTLD program, 
including:

http://forum.icann.org/lists/gtld-guide/msg00000.html
http://forum.icann.org/lists/gtld-guide/msg00026.html
http://forum.icann.org/lists/gtld-guide/msg00027.html
http://forum.icann.org/lists/gtld-guide/msg00046.html
http://forum.icann.org/lists/gtld-guide/msg00110.html

http://forum.icann.org/lists/2gtld-guide/msg00000.html
http://forum.icann.org/lists/2gtld-guide/msg00005.html
http://forum.icann.org/lists/2gtld-guide/msg00014.html
http://forum.icann.org/lists/2gtld-guide/msg00020.html

http://forum.icann.org/lists/irtp-draft-report/msg00000.html

2. For all the reasons we've previously discussed, we oppose the introduction 
of new gTLDs. We believe the process has been captured by a small group looking 
to directly make short-term profits from their introduction at the expense of 
the broader community through the resultant confusion for consumers and 
imposition of negative externalities upon 3rd parties. None of our comments 
should be taken as an endorsement of ICANN's plans or that of the IRT -- we are 
simply talking about degrees of damage if the plan in any form is adopted, the 
differences between an earthquake measuring 9.5 on the Richter scale vs. one 
measuring 9.4. Both are catastrophes where the costs far outweigh the benefits. 
It is incumbent upon the NTIA, DOC and DOJ to intervene in order to end the 
waste of resources associated with discussing a plan that the public has 
resoundingly rejected. 

3. While ICANN staff state:

http://forum.icann.org/lists/irtp-draft-report/msg00009.html

"The formation of the IRT does not preclude other contributions. Any member or 
set of community members, including other representatives of the IP community, 
is also welcome to include independent recommendations to be considered by the 
Board."

we believe this to be unsupported by facts. There is no evidence that the Board 
has considered contributions unless they have been heavily filtered by ICANN 
staff. The ICANN Board highly limits its participation in public mailing lists 
or other forums where the community can interact with them. Many in the 
community have made thoughtful and detailed recommendations that have received 
no consideration by the Board. Evidence of that consideration, if it existed, 
would be found in Board meeting minutes and in recorded votes.

4. By providing it with funding and promotion through ICANN's official website, 
ICANN is giving undue weight to the self-selected and unrepresentative members 
of the IRT. This is almost by definition a classic case of co-optation, the 
tactic of neutralizing the opposition of a group by assimilating them. We note 
that ICANN has not provided similar levels of funding to other thoughtful 
opponents of new gTLDs (ones who are harder to co-opt) to finance their own 
independent outreach or research endeavours. The unrealistic time period for 
comments further demonstrates that ICANN has already made its plans and is 
operating in a top-down manner to rush through an unsupported program over the 
objections of the community. This is in opposition to the bottom-up consensus 
driven manner that it should be operating. The secrecy with which the IRT has 
operated, without public mailing list archives and without MP3 
recordings/transcripts of meetings, violates the ICANN
 principle of transparency.

Indeed, the report stated (page 4 of the PDF):

"The work of the IRT is not endorsed by any particular IRT member, their 
clients, companies or affiliated companies or organizations, and participation 
in the IRT of any individual member does not signify endorsement, consent or 
approval of the work product of the IRT by any business, company, affiliate, 
client, association, group or any other party with which the individual IRT 
member may be affiliated or may otherwise represent. Indeed, any business, 
company, affiliate, client, association, group or any other party with which 
the individual IRT members may be affiliated or may otherwise represent, may 
disagree with or have differences with any one or several of the solutions 
proposed."

This needs to be kept in mind when properly weighing this report which is not 
even endorsed by its own members, against the weight of hundreds of thoughtful 
comments against new gTLDs by the world's largest and smallest companies, 
consumers, governments, and other organizations who were by far against any 
introduction of new gTLDs. To be more blunt, this report should be given the 
weight of a single feather, compared to the mountain of existing comments by 
stakeholders. Failure by the mountain to comment on the IRT's report should not 
be interpreted as acceptance of its proposals, but instead the IRT needs its 
own mountain of supporters to overcome existing opposition to new gTLDs.

5.There is a conflict of interest in IP lawyers and national Intellectual 
Property offices to be recommending solutions that will benefit their industry 
through resultant greater volumes of TM registrations worldwide (e.g. to meet 
criteria such as 200 trademark registrations in 90 countries, or having TM 
registrations for sunrise purposes, etc.). Thus, their interests are not 
necessarily aligned with their clients or other users of the internet and 
domain name system. If beef producers were asked for input on the DAG, they'd 
probably be suggesting solutions that would involve eating a lot of Big Macs 
and Whoppers in order to get a new gTLD. While we do have a list of IRT members 
that has been made public, we do not have a "Statement of Interest" from each 
of them that has been posted (at least I could not find one) as would be 
required in a GNSO task force or work group.

6. We oppose any recommendations that will impact existing gTLDs through the 
"equitable treatment" clauses in current contracts. e.g. for dot-com, the 
relevant clause is:

http://www.icann.org/en/tlds/agreements/verisign/registry-agmt-com-01mar06.htm

"3.2.(b)   Equitable Treatment.  ICANN shall not apply standards, policies, 
procedures or practices arbitrarily, unjustifiably, or inequitably and shall 
not single out Registry Operator for disparate treatment unless justified by 
substantial and reasonable cause."

bad policy choices in new gTLDs (including but not limited to the elimination 
of pricing caps) will propagate back into current gTLDs. Unfortunately while 
the IRT says it is for new gTLDs alone, the fact is that they will affect 
registrants who want nothing to do with new gTLDs. The IRT shouldn't be used to 
rewrite existing contracts through this backdoor, but should instead go through 
the GNSO Consensus PDP route.

II. Issues with the IRT Draft Report

1. In general, the report makes no attempt to perform any cost-benefit analysis 
for its recommendations, even in rough or qualitative terms (as opposed to 
precise quantitative terms). From the list of team members, we see few with 
economic or business backgrounds, but instead the team is filled with IP 
lawyers. While lawyers have their place in helping to create precise legal 
language, it also means that solutions that are superior in economic efficiency 
to those proposed by the team were not even considered, simply due to the 
biased manner in which the team was self-selected.

2. IP Clearinghouse Proposal

a) We oppose ICANN funding for any "IP Clearinghouse." ICANN funding ultimately 
comes from domain name registrants, meaning higher costs imposed upon them. 
Domain registrants do not want to be funding a $100 million+ per year 
organization.  ICANN needs to be a lean and focused organization, and should be 
shrinking in size, not one that is engaging in mission creep to create and 
operate databases outside its scope of expertise. 

b) The description of the IP Clearinghouse is so broad that it need not even be 
applied to domain names (why not use it for vanity license plates, vanity phone 
numbers, etc.). It even mentions "forerunners of the IP Clearinghouse" that 
apparently already exist. If these already do exist, ICANN can mandate that 
registry operators use them, but ICANN should not be the one to either create 
or pay for them. That would be another boondoggle, in our opinion, a waste of 
the money that hardworking registrants pay into ICANN.

c) Any IP Clearinghouse can be created and funded by TM holders, national 
intellectual property offices, or even the private sector, without ICANN's 
involvement. For example, DomainTools.com has a "Watch Service" for domain 
monitoring. I am certain they are not alone. ICANN can mandate that the data be 
provided by registry operators to the public at cost or free (e.g. zone files, 
etc.) to allow the private sector to create that solution. Those private 
competitors can create their own clearinghouses and deal with registry 
operators and TM holders on their own.

d) We do not need another set of "contracted parties", the "IP Clearinghouse 
Providers" in the ICANN organizational chart, to be lobbying for extra money 
and resources.

e) More cynically, these "IP Clearinghouse Providers" have a conflict of 
interest, as more new gTLDs mean more business for them (see the "feather" 
argument above when weighing their input, or the "parasitic consultants" in 
previously submitted comments).

f) Furthermore, this "IP Clearinghouse" as described by the IRT is yet another 
"monopoly provider" (as opposed to multiple operators run entirely by the 
private sector and outside of ICANN's scope). Monopolies are a bad thing for 
the most part and should be avoided where possible.

g) We oppose the list of "supported applications" of the IP clearinghouse being 
anything other than informational in nature, in particular the use as a 
"uniform rapid suspension system" or as a "globally protected marks list." 
(more on these below)

h) The IP Clearinghouse is yet another failed technological solution to an 
economic problem. The myth behind it is that if we build some monolithic 
technological solution (a big database), all our problems will be solved. This 
is simply not true. Watch services *already exist* -- have they caused 
cybersquatting to disappear? Nope. Will having an even "bigger" watch service 
magically solve the problem now? Of course not. [just like the entire premise 
of new gTLDs --- launching .biz, .info, .pro, etc. didn't lower dot-com costs 
for consumers; ICANN's "solution" is to follow up that failure with even *more* 
new gTLDs; ICANN is like a dieter who failed to lose weight when they bought a 
"miracle" electric ab stimulator. After failing to lose any weight wearing it 
for 1 hour per day, their "solution" is to buy 10 ab stimulators and wear them 
for 10 hours/day, instead of changing their approach completely to eat less and 
exercise more.]

Major lawsuits have been launched at some of the biggest cybersquatters. Has 
this solved the problem? Not at all. The folks who've been gaming the system 
will continue to do so, unless they are wiped out economically. New gTLDs 
simply create new playgrounds for them to play new games. The solution should 
be to eliminate the idea of any new playgrounds until existing ones are cleaned 
up first. Wipe them out of business. Have class-action lawsuits that wipe them 
out (the single-plaintiff ones have been ineffective, obviously). It's a poor 
military strategist who would willingly go into a war with multiple fronts. 
Each new gTLD represents a brand new front. Win the war in dot-com (and other 
existing TLDs) first, before thinking of opening up new fronts that simply 
divide your attention and resources.

i) Clearinghouses do not easily take into account visual or aural similarity, 
or how marks are used. Ultimately the database solution simply doesn't scale. 
Human involvement will always be necessary. If clearinghouses scaled, and 
technological solutions scaled, then UDRPs would be handled by  an automated 
computer, instead of by a group of human beings.

j) We oppose all the "principles" (page 10 of the PDF) behind the operation of 
the IP Clearinghouse. It's simply something that should be left to the private 
sector, instead of being a contracted party with ICANN. Given ICANN's history 
working with overpriced contractors, we reiterate our opposition to more 
monopoly boondoggles that will be ultimately paid for by dot-com registrants 
(who constitute the vast majority of existing and future registrants, let's be 
honest), and not by new gTLD users themselves.

3. The Globally Protected Marks List

a) We oppose the introduction of any list that gives favourable treatment of 
one rightsholder relative to another, and in particular TM rights relative to 
other kinds of rights. In particular, from our past comments highlighting the 
views of Tim Berners-Lee on new TLDs being harmful:

http://forum.icann.org/lists/2gtld-guide/msg00020.html

"The value of each domain name such as example.com also drops because of brand 
dilution and public confusion. Even though most people largely ignore the last 
segment of the name, when it is actually used to distinguish between different 
owners, this increases the mental effort required to remember which company has 
which top level domain. This makes the whole name space less usable."

Note, that not only are so-called "famous" trademarked domains diminished in 
value. It happens to EACH and EVERY domain name. 

The GPML, on the other hand, is saying that one set of rules applies to 
everyone, EXCEPT certain "special" people. Everyone of course likes to think of 
themselves as "special" just as a parent thinks their newborn baby is the most 
beautiful baby in the world.

b) The GPML is an attempt to create new law through arbitrary and self-selected 
criteria of a small secret group. This is also a form of co-optation, as 
described earlier, where a few large companies whose TMs fit the criteria and 
who oppose new gTLDs currently are induced to "hold their nose and say nothing" 
by getting special privileges, at the expense of everyone else (smaller 
competitors, the public, consumers), suffering.

For example, I've not looked it up but a famous mark like Ralph Lauren's "Polo" 
might fit the criteria. It might not, but let's say for example it did fit. It 
would be granted special rights relative to those who are simply interested in 
the sport of polo (horse or water polo).

c) The GPML disadvantages smaller companies and individuals unduly. It's like 
saying that the depreciation rate on computers should be 100% if you're IBM, 
but if you're a startup company, your depreciation rate is only 30%. It's this 
kind of lobbying by special interest groups that overcomplicates regulations 
and creates unfairness. Creating a GPML is an attempt to solve one "wrong" (new 
gTLDs) with another wrong (GPML), and two wrongs simply do not make a "right", 
as any parent will tell a child.

4. Protection for GPMS at the top level

a) Obviously since we do not support the idea of a GPML, we do not support any 
special protection for them at the top level.

b) We DO agree with the IRT that the String Similarity Algorithm is extremely 
flawed, and needs to take into account other factors as per our prior comments, 
including aural, braille, meaning, and other forms of similarity (based on 
future user input technologies).

5. Watch Notice for Top-Level Domains

a) While a watch notice might seem to have advantages, it is simply another 
boondoggle that we cannot support, as we don't support the IP Clearinghouse as 
discussed above. 

b) Instead, we believe that any new gTLD applicant should be compelled to give 
wide public notice of their applications (e.g. placements in the Wall Street 
Journal and other papers of record world-wide).

http://en.wikipedia.org/wiki/Newspaper_of_record

Such a list of newspapers of records that cover at least the G20 countries and 
preferably more (e.g. The Economist Magazine, etc.) should be something that I 
believe the IRT would support. 

c) ICANN can provide application details in XML and as RSS feeds. Commercial 
and private companies would be free to parse and collect the data from ICANN's 
website and use it for their own providers.

d) The combination of (b) and (c) would not only inform more stakeholders, but 
would do so far more economically than the creation of an IP Clearinghouse 
monolith.

6. IP Claims Service

a) While the IP Claims Service solution might appear attractive at first 
glance, it is clear we are simply going down the road of past failures, and not 
learning from history. If the idea (or "myth") behind new gTLDs is that they 
are going to expand the namespace, then how do these RPMs accomplish this? 
Honestly, they simply do not work. To go back to the prior example of Ralph 
Lauren's "Polo" mark, if an application for .sport was approved, why should 
Ralph Lauren have superior rights to polo.sport relative to anyone else, when 
it would simply duplicate their polo.com and not expand the namespace? The 
principle of first-come first served should be preserved, if the IRT is going 
to be supportive of new gTLDs. Obviously the even better solution is to have no 
new gTLDs at all, but if the IRT is going to be supportive of them, then they 
can't be bought off for that support by undermining the supposed reason for the 
new gTLDs, namely expansion of the
 namespace. 

b) The IRT should be compelled to make a clear statement as to whether or not 
they support expansion of the namespace (to followup on point (a) above), 
because if they are supportive of the expansion, they need to justify why their 
"solution" is to diminish that expansion and simply reduplicate dot-com for 
their brands.

c) The notion that for GPMs that there should be a "be a process by which these 
applicants should be permitted to register the initially-blocked names if they 
can demonstrate to an independent third party that their registration of those 
names would be consistent with generally accepted trademark laws; namely that 
its use of the domain name would not infringe the legal rights of the GPM 
owner" is a reversal of trademark law, and reverses the presumption of 
innocence. Folks are assumed guilty if they want to register polo.sport, or 
pontiac.tribes, and need to prove they are innocent.

d) The procedure in relation to non-GPM marks simply repeats failed solutions 
of the past. The reason we have cybersquatting isn't because people are unaware 
of what they're doing --- the vast majority of it, especially amongst the 
"industrial cybersquatters" know exactly what they're doing. This 
"affirmatively opts into" language is simply another way of rewriting the RAA 
agreement, or modifying it without a consensus process as required within the 
GNSO structure of ICANN.

7) Standard Sunrise Registration Process

a) We reiterate our comments of 6.b) above by reference. If the IRT truly 
supports expansion of the namespace, than how does reducing its space through 
re-duplication of .com accomplish that objective? Of course, one can eliminate 
the hypocrisy by opposing new gTLDs as our company does. 

b) Any sunrise period is simply a way for the registry operator to price 
discriminate, i.e. to charge higher prices for the "desirable" domains, i.e. 
the 100,000 common words, acronyms, brand names, etc. that are in demand for 
SEO purposes, vanity, and defensive registrations. Let's be honest -- this is 
where registry operators want to milk their new gTLD as much as possible.

c) Once again, the IRT is recycling solutions that have failed in the past. The 
only new ideas are a new set of acronyms like "SER" or "SRDP" that they hope to 
add to the ICANN glossary --- this is hardly anything worth congratulating.
 
8) Potential Gaming of TM Applications

a) The entire IRT report implicitly assumes that the TM database itself has not 
been gamed. Obviously, this is false. As seen in multiple sunrise periods, 
sophisticated individuals and organizations will seek out trademark 
registrations for popular terms well in advance of any sunrise (i.e. and are 
currently in effect). For instance, in the EU ADR decision for "AUTOTRADER" it 
was asserted:

http://www.adreu.eurid.eu/adr/decisions/print.php?dispute_id=3147

"On 22 November 2005 the Respondent applied to register a Benelux Trade Mark 
for Auto Trader in Class 15 of the Nice Classification system in respect of 
“plectrums” a category of goods unrelated to the natural meaning ascribed to 
the words Auto Trader."

"The Complainant submits that in addition the Respondent has filed over 100 
trade mark applications in the Benelux Registry all of which were registered on 
22 November 2005.  The trade marks were registered by the Respondent and 
another company, First Internet Technology Limited that is registered at the 
same address.  The trade marks registered by the Respondent include marks such 
as “Digital  Camera”, “Mortgage Loan” and “Online Poker” for goods and services 
such as “computer games for Catholic kids”, “services of a funeral director” 
and “table silver” respectively."

And in the EU ADR decision for "ASK" it was asserted:

http://www.adreu.eurid.eu/adr/decisions/print.php?dispute_id=2438

"The Panel quite frankly does not believe the Respondent's submission as to the 
true intention behind the Benelux Mark.  In the Panel's view, the more credible 
explanation of the Respondent's interest is offered by the Complainant who 
points out that the Benelux Mark was filed purely for the purpose of 
application for the disputed domain name.  The terms of the Respondent's other 
Benelux trade marks, of which some 47 were listed by the Respondent following a 
request by the Panel, are apparently very similar; in general they consist of 
an English word for which the associated goods and services are generally 
plectrums; all of which were applied for on 22 November 2005."

These are not isolated incidents. There is rampant abuse of TM applications for 
the sole purpose of capturing domains in sunrise periods (and even for 
capturing gTLDs themselves), and the IRT report fails to recognize or discuss 
this. By proposing special rights for TM holders, it encourages gaming of TMs 
themselves. For every action, there will be a reaction. The IRT members 
implicitly make the assumption that the folks whose behaviour they wish to 
discourage are unsophisticated and poor, when the opposite is true. Indeed, 
some of the most malevolent and well funded might engage in acquiring TLDs of 
their own.

b) In addition to the types of abuse in (a) other trademark registrations have 
been made in order to reverse-hijack the domain names owned by others. The IRT 
does not discuss how they will prevent this sort of gaming either.

9) Uniform Rapid Suspension System (URS)

a) We oppose the introduction of the URS as drafted by the IRT team, as it is a 
violation of due process. While there are good intentions behind it, it will 
lead to abuses.

b) The facts are that the number of UDRPs in relation to the total number of 
domain names is extremely small (say 5,000 UDRPs per year divided by 100 
million domain names). That's a rate of 5 per 100,000 per year, roughly 
speaking. That's below the homicide rate in the USA in 2007 of 5.9 per 100,000:

http://www.infoplease.com/ipa/A0873729.html

We agree, though, that the amount of abuse is rising, however, the URS is not 
the solution. I'd invite the IRT to participate in the Registration Abuse 
Policies working group of the GNSO instead:

http://forum.icann.org/lists/gnso-rap-dt/

c) The more appropriate method to deter abuse (not only TM abuse, but other 
abuses) is through WHOIS verification, which we've advocated for a long time. 
See our prior comments in:

http://forum.icann.org/lists/gtld-guide/msg00026.html (F, Module 5, point 15)

"To deter abuse, we suggest that there be a system of address verification by 
registries in new gTLDs, similar to what exists in some current ccTLDs, in 
order to stem the tide of domain name abuse. No domain would resolve until a 
mailed PIN code was entered into a central system, so that the registrant is 
verified. This would ensure a higher degree of registrant WHOIS accuracy, 
reducing abuse from those who routinely use fake WHOIS. This should eventually 
be adopted into existing gTLDs as a universal standard, with financial 
penalties to registrars who permit fake registrations above a certain level."

We've advocated the same position in the fast flux workgroup:

http://forum.icann.org/lists/gnso-ff-pdp-may08/msg00410.html
http://forum.icann.org/lists/gnso-ff-pdp-may08/msg00638.html
http://forum.icann.org/lists/gnso-ff-pdp-may08/msg00643.html
http://forum.icann.org/lists/gnso-ff-pdp-may08/msg00730.html
http://forum.icann.org/lists/gnso-ff-pdp-may08/msg00870.html
http://forum.icann.org/lists/gnso-ff-pdp-may08/msg00893.html

and elsewhere. This allows private parties to seek private solutions to their 
problems, without making ICANN into the universal police of the internet.

d) We sympathize with the fact that many brand owners spend large amounts of 
money to build up portfolios of domain names they do not want (i.e. defensive 
registrations). Indeed, this would be a sweet spot for new gTLD registry 
operators, the focus of many of their business models. The solution isn't to 
have a URS, but instead to reduce the price of these unwanted domains.

i) First, the IRT should explicitly support price caps for all new gTLDs (and 
existing ones too), and also that registry operators for a gTLD be selected 
through competitive tenders for fixed periods, just like any other procurement 
contract. This is aligned with the NTIA/DOJ/DOC recommendations we've discussed 
and supported before.

http://www.ntia.doc.gov/comments/2008/ICANN_081218.pdf

ii) The IRT should support a pricing structure for domains that consists of 2 
components, namely a price for the registration itself, and a separate 
(optional) price to have the domain name resolve (i.e. to have nameservers 
associated with that domain name). Thus, a domain name that is registered 
purely for defensive reasons that doesn't need nameservers (and thus would not 
resolve) would have a lower carrying cost. For instance, if we assume that 
under a competitive tender situation .sport prices are $3/yr under the current 
system, that might be $1/yr for the actual registration and $2/yr for insertion 
into the zone file. By separating out the fees, names without nameservers that 
don't resolve would be considerably less expensive. Also, by separating out the 
fees, there could be separate tenders for the DNS resolution of a zone and the 
management of the zone itself.

For example, .ca and many other ccTLDs might run some of their nameservers 
themselves, but also supplement those nameservers by ones operated by UltraDNS 
or other companies. This is similar to how the root itself operates, with 
ICANN/NTIA/DOC managing the master root zone file itself, but the nameservers 
(resolution of the zone) are handled not only by them, but by other companies, 
universities, the US military, and other organizations.

By breaking this into 2 components, we would receive economic benefits, and the 
TM benefits (lower defensive registration costs) across all gTLDs.

e) Any balanced URS, if one were to exist (which we don't support as drafted by 
the IRT) would take into account the age of a domain, and allow registrants to 
white-list themselves (perhaps via verified WHOIS as we discussed above). A 
domain name registrant who has owned their domain for 10 years should be able 
to take a month long vacation without worrying that they might lose their 
domain by not responding to some email notice that they've never actually 
received, and should be given ample time to obtain legal counsel to defend 
their property. This puts the onus on the mark holder to not wait years before 
asserting a claim. While not going so far as to requiring laches or statutes of 
limitations to bar a claim, domain holders are entitled to reasonable amounts 
of time to respond, determined by law and not by those making allegations. One 
formula might be:

Time To Respond = Age of domain (in months) + 15 days

Thus, for a freshly registered domain (0 months), one would have 15 days to 
respond to a complaint. A 2 year old domain (24 months) would have 39 days. A 
10 year old domain (120 days) would have 135 days.

f) The draft URS is essentially a way of rewriting the UDRP without going 
through the PDP process, and on that basis we oppose it. It would also affect 
registrants in existing gTLDs via the "equitable treatment" clause as discussed 
above.

g) Allowing complainants to go forward without posting a security bond is not a 
"complication" as the footnote on the bottom of page 22 of the PDF declares, 
but is a reasonable step to prevent abuse. URS complainants are asking for an 
extraordinary action to take place, one that could cost an innocent domain 
registrant millions of dollars in damages (e.g. loss of search engine position, 
loss of ecommerce sales, etc.). If the complainants are not going to be posting 
a bond, then the decision provider should explicitly be liable for damage and 
be compelled to carry $100 million in liability insurance, and also to 
explicitly consent to the jurisdiction of the registrant. This might cause 
problems for WIPO, which does not like to be dragged into court (and probably 
asserts that they should not submit to any jurisdiction at all). We've seen 
some frivolous decisions in the UDRP process (that later got overturned in real 
court). UDRP panels would be more
 responsible and accountable if they were on the hook financially (or their 
insurance rates would be higher if they hired sloppy panelists) for incorrect 
decisions that caused damage to registrants. The URS is simply a lower cost and 
expedited system to cause damage through false positives (i.e. claimed abuse 
that does not actually exist, but that results in damage to innocent 
registrants).

h) We do not believe registrars will accept the indemnifications of 
complainants (as described in page 22 of the PDF), and could be held liable for 
incorrectly taking down a domain of their customers. Indeed, many reputable 
registrars (we use Tucows/OpenSRS, a highly reputable one) KNOW their clients, 
and are in a far better position to make a decision than a URS case manager. 
Registrars should not be compelled to act against the interests of their own 
clients if they actually have superior knowledge of that client.

i) We've also suggested in the past that WHOIS be supplemented by the use of a 
"Legal" contact. This Legal contact can be used for notices of disputes 
(including the URS), to ensure actual notice and reduce default decisions. This 
Legal contact could either be visible in the WHOIS output, or optionally 
visible only to the registrar.

j) Obviously notice via email is insufficient and unreliable. We found it 
ironic and hypocritical that WIPO, in discussing their own proposals/solutions:

http://www.icann.org/correspondence/wilbers-to-pritz-03apr09-en.pdf

sent documents to ICANN via both email *AND courier.* Even the registered 
letter sent 5 days after the complaint is filed (which can be really 7 or 8 
with weekends and holidays) might not be received until after 14 days has 
passed. Thus, it is highly probably that actual notice will never be received, 
or leave insufficient time to contact a lawyer, etc.. No attempt would even be 
made to send a complaint via FAX when a registrant might have good FAX info in 
the WHOIS, or to other contacts besides the registrant (i.e. the admin contact, 
the technical contact, the billing contact can all be different than the 
registrant). Obviously this is unacceptable. I would imagine half the IRT 
members would not be able to provide a response within 14 calendar days of a 
complaint if an attacker (and I do consider abusive complaints to be attacks) 
carefully crafted the complaint date (i.e. around Xmas holidays, etc.). With 
the prevalence of spam, email is simply
 unreliable, and compound that with vacations, life beyond monitoring domains, 
etc., the short notice period is unacceptable.

k) As above, registrants should be able to opt-out of any URS (and also the 
UDRP), via whitelisting themselves (e.g. having valid legal contact, verified 
WHOIS, or other steps to indicate they are not abusive).

l) Just as in the UDRP, the registrant is unable to do discovery or cross 
examination of the complainant in order to contradict their evidence. This can 
lead to a contest of who can fabricate better "evidence" (e.g. by manipulating 
images in Photoshop or other procedures).

m) IRT team members can/should consider approaching the upstream providers of 
domain monetization or hosting for relief (e.g. Google/Yahoo are the major 
pay-per-click providers), as well as webhosting companies and DNS providers. 
Registrars are not the only means to suspend an abusive domain.

n) Footnote 12 notes that the fee level would be in the area of $35 to $40. My 
company owns extremely valuable domain names (as I'm sure do many IRT members 
and their associated companies). How much time do IRT members actually believe 
examiners will spend on a complaint when they're only making $20 or $30 
(assuming the provider takes a cut), and less after taxes? This must be some 
kind of a joke, that complaints will be handled by the WIPO equivalent of 
burger flippers. Or, perhaps examiners will be outsourced to poorer countries 
like India or China where $20 is a lot of money, but which might lack the 
qualifications and experience of those in first-world economies. Upstanding and 
responsible companies like ourselves should be able to pay $50, then, on 
deposit with our registrar, as a "security bond" to whitelist all our domains, 
to say "no" to having cases handled by McPanelists using URS as a McJob (with 
apologies to McDonald's).

o) The "default procedures" (page 26) are clearly one sided in favour of the 
complainant. First off, as above they don't allow the registrar latitude to 
come to their own decision, if they actually have superior knowledge. Second, 
they certainly don't even allow the examiner to look more closely into a matter 
(beyond the "prima facie" level). Thirdly, even a defective answer is 
completely ignored, even if it was filed in a timely manner (with the terribly 
short time limits which we objected to). Certainly UDRP history has taught us 
that average responses often make mistakes (and need to be amended). We are not 
even shown sample forms for Answers, to determine how much prone they are to 
incomplete answers by innocent respondents making casual errors (because they 
didn't have time to get a lawyer, for example), relative to pro se UDRP cases.

p) The IRT doesn't appear to understand how DNS works, as evidenced by their 
statement that the "domain name will be deactivated from use in connection with 
any registrant's website. Instead, a standard page will be posted in connection 
with the domain name...."

Most registrants do not use the nameservers of their registrars (i.e. they use 
that of their webhosting companies, etc.). Thus, in order to post a "standard 
page", i.e. create an A record associated with a website under the control of 
the registrar, the registrar will have to change the nameservers to their own 
control. This means that if the domain name was being used for any other 
purpose besides the web, e.g. email for 10,000 people, DNS services for 500 
other domains on a shared hosting service, game servers, FTP servers, backup 
servers, time servers, they would all immediately fail. The internet is more 
than just the "web", and the URS doesn't appear to understand this basic fact. 
The amount of collateral damage can be significant from wrongly deactivating a 
domain name, because the URS doesn't preserve the "status quo" --- thus the 
consequences from an incorrect URS decision are significantly more dangerous 
than the UDRP (which does preserve the
 nameservers, pending appeal), yet the decisions are being handled by 
McPanelists as McJobs as noted in point n above. 

q) The IRT goes on to say that "During the period of default, the Registrant 
cannot (a) change the content found on the site...." Since the site could have 
been hosted on a server not controlled by the registrar, the registrant could 
certainly change that content to anything they like. It would just not be 
visible to anyone. If there was legitimate content on it, and did not need to 
be changed, it would also not be visible (and this would make a Default Answer 
likely fail, as the examiner would not be able to independently see what the 
website looked like before it was switched to the nameservers of the registrar).

r) The IRT report says that "To avoid improper gaming of the system, the 
decision of the
URS Examiner shall be based on the status and use of the disputed domain name at
the time of the filing of the complaint." We certainly doubt the ability of the 
panelists to create an active snapshot of ALL uses of a domain (every single 
webpage, not just the homepage, subdomains, email usage, FTP usage, use as 
nameserver for other domains, etc.) and all for $40. Instead, we expect that 
this "status" will simply be what the web homepage (if it even exists) will be 
archived, a single page demonstrating a possibly distorted and biased use of 
what could be a very complex domain. For instance, Google has in its index 535 
million pages for the Yahoo.com domain name (doing a "site:yahoo.com" search at 
Google). If someone filed a URS against Yahoo, which of those 535 million pages 
will the examiner be expected to look at, at their $20 McJob?

s) Page 27 shows that the IRT has been constructed in an unrepresentative and 
twisted fashion, and appears to be a cut/paste job of WIPO's April 3rd letter 
to ICANN. In particular the 3rd element under examination is stated as:

"Whether the domain name has been registered ** OR ** used in bad faith." 
(emphasis added)

This is different from the UDRP, which requires BOTH elements, i.e. "AND". Aha! 
Gotcha! Bad boys and girls. This is clearly an unacceptable revision of the 
UDRP, a backdoor way of rewriting it in order to make it esaier to reverse 
hijack domains.

t) Page 27 also suggests that trademarks need to be issued by a jurisdiction 
that conducts substantive examination of a trademark application prior to 
registration. This is clearly something even the best jurisdictions advertise, 
but in reality many are sloppy. For example, had it not been for our own TM 
Opposition, the Canadian Intellectual Property Office would have gladly issued 
a TM for "bling" in the field of jewelry. See:

http://cipo.gc.ca/

(click on trade-marks database, search for "bling" and go to page 2, 
application 1234016) 

Of course, "bling" is a dictionary term directly related to jewelry, that long 
predated that application, and our lawyers crushed that application once we 
opposed it (the applicant abandoned the application completely).

http://www.merriam-webster.com/dictionary/bling

Preposterous TM applications squeak through the system all the time, and if 
they're not caught before registration, are challenged after registration when 
TM holders try to assert them. There are trademark trolls

http://en.wikipedia.org/wiki/Trademark_troll

who would certainly abuse the URS just as they do the UDRP, and the URS would 
embolden them as currently drafted (it certainly lowers their fees and makes it 
easier to get submarine default decisions). There are UDRP decisions that have 
gone the wrong way, in our opinion, which embolden those kinds of reverse 
hijacking attempts.

u) Only a year penalty for filing *3* abusive complaints? Certainly you're 
joking. The bar must be set much higher, folks. This is why there should be 
security bonds or insurance policies, to ensure that registrants who are abused 
can receive justice, as certainly they cannot rely upon the safeguards proposed 
by the IRT to date.

v) Obviously the registrant should not suffer from deactivation of their domain 
name until they've had an opportunity to appeal in court (even the UDRP allows 
this), or go to court instead of the URS (i.e. override by going immediately to 
court, and ignoring the URS). Often that's the optimal response to a UDRP, to 
demonstrate that one will not be abused by aggressive complainants.

10) Post-Delegation Dispute Mechanism

a) At this point, the horse has left the barn. The focus should be on 
eliminating bad operators before they get their hands on a gTLD in the first 
place. Not granting new gTLDs in the first place is the simplest approach, 
given the vast opposition. The other "obvious" point is to not give presumptive 
renewal to registry operators, or give them unlimited pricing power. Those are 
recipes to attract abusers.

In our view, it should simply go to court, as at this stage ICANN has already 
failed. There's no need to create new projects for WIPO and other dispute 
providers. The Damocles Sword of litigation, combined with large security bonds 
(or insurance policies) posted by registry operators would mitigate some of 
their potential incentives to not follow their own stated policies. But, it's 
so subjective and open to gaming that many will simply factor in legal expenses 
as their cost of doing business, especially when their real intention is to 
profit from sunrise periods, landrushes and uncapped prices in a short-term 
manner, and then walk away with their booty. 

11) Thick WHOIS and Universal WHOIS

a) We support thick WHOIS, and furthermore registrant address verification via 
PIN codes (as discussed above and in prior comments). Domains can be 
registered, but not resolve until the registrant passes verification of their 
WHOIS.

b) We find it highly amusing that the IRT is only now appreciating the 
importance of Universal WHOIS. Where were they when myself and others were 
inquiring about it since 2003, in relation to VeriSign's Appendix W obligations 
under their prior contract (an obligation that they negotiated away)? 

http://www.icann.org/en/tlds/agreements/verisign/registry-agmt-appw-net-org-16apr01.htm
http://gnso.icann.org/mailing-lists/archives/ga/msg00523.html
http://gnso.icann.org/mailing-lists/archives/ga/msg00822.html
http://gnso.icann.org/mailing-lists/archives/ga/msg01690.html
http://gnso.icann.org/mailing-lists/archives/ga/msg03065.html

as just a few examples. Do a search for: kirikos "universal whois" (with the 
last 2 words in quotes) or for: kirikos "appendix w"  and you'll find more 
examples. If ICANN couldn't compel VeriSign to produce Universal WHOIS even 
though VeriSign was supposed to spend $200 million in R&D and deliver it 
substantially by the end of 2002, what makes anyone think ICANN will do it now? 
This is something that IRT members need to complain to the DOC, DOJ and NTIA 
more vigorously about, as they were certainly asleep for the past 6 years. 
ICANN never even posted the alleged R&D reports documenting VeriSign's 
expenditures. Universal WHOIS should be a precondition to ICANN getting JPA 
renewal, let alone proceeding with the gTLD boondoggle. New gTLDs need to wait 
until ICANN gets its house in order. 

The money for Universal WHOIS should come from VeriSign and other gTLD 
operators, with no additional fees paid by domain registrants (or money coming 
into ICANN from registrants). This was something that should have been 
delivered long ago. It should also not be "maintained" by ICANN, but should be 
put out to competitive tender and *overseen* by ICANN. ICANN has not proven it 
can operate any scalable database, let alone WHOIS. Leave it to experts, and 
send the bills to the registry operators.

12) String Confusion Algorithm

a) This algorithm was a complete was of money. We reiterate our suggestion in 
5.b above, namely that applications be widely advertised in newspapers of 
record (e.g. the Wall Street Journal, The Economist and other global 
publications).

III. Other Solutions

1. No Wildcarding

The IRT has not mentioned that wildcarding of new gTLDs (and existing ones) 
should be explicitly barred. This is a major source of typosquatting as folks 
are aware in .cm (Cameroon) and other TLDs.

2. Ascended TLDs

We propose (in a separate document, see separate comment in this comment 
archive) that instead of using the convoluted approach that ICANN has taken, 
that if new gTLDs are to be introduced (and we continue to oppose their 
introduction) that this be done using an entirely different method than they've 
taken in their DAG, one that is economically superior and would be more 
consistent with ICANN's mission of stability. I believe it would also be less 
offensive than the current DAG to members of the IRT, as it handles TM issues 
in a more elegant manner. See that separate document for a description, as this 
comment is long enough.

IV. Conclusion

In conclusion, the IRT while well-intentioned has produced no magic solution 
that paves the way for new gTLDs to be introduced. Their report has no easy 
fix. The IRT should be disbanded, and GNSO work groups should be formed to 
study the issues raised over the coming years, to achieve a bottom up consensus 
with the involvement of many stakeholders including those not already engaged. 
Greater outreach is necessary. ICANN should suspend further plans to produce 
guidebooks, and focus on more pressing matters like DNSSEC, IPv6 and shoring up 
the security of existing gTLDs.

We would be happy to answer any questions the IRT might have on our submission 
(feel free to contact us via email from our website).

Sincerely,

George Kirikos
http://www.leap.com/



<<< Chronological Index >>>    <<< Thread Index >>>