ICANN Logo

Report of the Internationalized Domain Names Working Group–Responses to Survey B

Posted: 28 August 2001




Appendix—Responses to Survey B: Policy Questions

1. What is your view of the value of IDNs? Who will benefit from them? Is there any empirical proof of these benefits? Who will IDNs harm?

AIPLA In the AIPLA's view, IDNs do have a value. Internet users for whom English is not their first language will benefit by being enabled to register and use domain names in their native languages.On the other hand, the introduction of IDNs will increase the likelihood of cybersquatting, in that foreign equivalents of Roman-character marks will be available for registration on the Internet.
WALID The significant value of IDNs is to make the Internet - for social, educational, communications and commercial purposes -- more fully accessible to the 92% of the world's population that does not speak English. If one can measure potential benefits based on demand, the early experience of the VeriSign GRS multilingual testbed demonstrates that there is a large and immediate need for IDNs, with VeriSign having registered some 920,000 IDNs in the first five months of operation.
Verisign Internet use is increasing dramatically throughout the world. Users who speak a language other than English comprise one of the fastest growing groups. Current estimates are that non-English speakers will make up two-thirds of all Internet users by 2003. Despite this growth, the Internet remains an English-centric resource with barriers of entry to non-English speakers. One such barrier is a lack of IDN capability within the Domain Name System (DNS). Currently, the DNS supports only domain names consisting of letters from the Roman alphabet and digits. In response to the demand for support for domain names in characters used by other languages, the VeriSign Global Registry Services has developed the IDN Domain Names Testbed. We believe that by making it easier for people around the world to use the Internet in their own languages, internationalized domain names (IDNs) constitute an important step towards making the Internet a truly global medium. VeriSign Global Registry Services (VeriSign GRS) is now able to offer the opportunity to register domain names in scripts that support languages used by approximately 80 percent of the world's population.
IPC The IPC is in favour of a stable IDN system. Users of non-Roman script in the world are in the majority and in the future this is likely to also be the case on the internet. Companies operating in countries in which non-Roman script is used wish to be able to use for their domain names the same designation for their company and their brands as they use in those countries. The ability to do this in a stable harmonized system will benefit local and international companies and is likely to assist local use of the internet.
Neteka

Neteka believes that the introduction of multilingual domain names is imminent and brings huge value to the Internet for it to become a truly global medium. Domain names have become more than merely a string of characters for technical identification of an Internet resource, and have become a critical part of corporate branding and more important of all the identity of a person or a corporation on the Internet. This is confirmed by the need for the UDRP and its importance.

The Internet community will benefit the most especially for the new comers on the Internet from regions where English is not a common language. The introduction of multilingual domain names reduces the barrier for entry for these people of the world. Corporations can also advertise their brand names in local languages, consistent to their existing literatures.

Register.com

The current DNS, which only supports the ASCII character set, creates a barrier to the use of the Internet for people who do not speak English or other Latin alphabet-based languages as their native language. The primary benefit to developing IDNs that are fully compatible with the current DNS would be to broaden the accessibility of the Internet, allowing more people to participate in and contribute to the Internet community.

According to IDG News Service, 47% of Internet-connected devices in the Asia Pacific region use written scripts that require the use of non-English character sets. While it is difficult to determine precisely how much IDNs will increase the use of the Internet by these communities, such indications of demand for multilingual products lead us to believe that IDNs will greatly encourage and simplify Internet use.

While the addition of new languages may cause difficulties for individuals who would like to access, for example, a Chinese language website but do not have the technical ability to type in Chinese, we at register.com feel that the benefits gained by this option to Chinese speakers considerably outweighs such disadvantages.

Peacenet

1-1. What is your view of the value of IDNs?

IDN could make people convenient for using Internet, especially those whoare using non-Latin family languages. Moreover, it would make an cultural effect on them so that they could exactly express their own identities.

1-2. Who will benefit from them?

IDN could be very beneficial to those users who are using non-Latin family languages and those entrepreneurs that are trying to make some businesses targeted on them.

1-3. Is there any empirical proof of these benefits?

As IDN has not yet been deployed, its benefit is very uncertain. Then, in other services like BBS or on-line chatting, where user's mother tongue could be available in using id there are lots of people who prefer to choose their mother tongue style ids. This shows us well how such a use could be beneficial.

1-4. Who will IDNs harm?

If we suppose a situation, in which IDN is to be popularly used and so the number of users who use only IDNs would sharply increase, the communication problem among people or companies that belong to different language zones could take place. In these cases, those who could not key in specified different language domain name labeled on the printed matter could fail to communicate or give up such attempts.

JPNIC

IDN is very valuable because the name space is broadened for use by non-English users. This is because

  • Many Japanese words have multiple presentations, i.e., in Katakana, Hiragana, or Kanji.
  • Different Japanese words usually have the same presentation in ASCII string, as they have the same pronunciation or they have the same corresponding English words.

We have small set of examples by now as registration has just begun. Our experience till now shows more than half of the applied domain names in the sunrise period of a new domain space, which accommodates both ASCCI and Japanese names, were Japanese ones that are trademarks or trade names. This is thought to be a kind of evidence that Japanese users would like to use Japanese domain names.

However, visually handicapped users may suffer from the difficulty in identifying the domain names they want to type-in, because pronouncing English alphabets is much easier than vocally identifying Japanese characters among over 2,000 different characters.

TWNIC

It has been noted that the number of non-English speakers is increasing very quickly in the whole Internet Community. However, DNS, which was basically designed for English speakers, is significantly unfriendly to the non-English speakers. Since the Internet is the medium for communication, it should never be used as a test on people's English proficiency.

In China, over 80% of the population cannot read English (why should every body learn English?). Even though extensive Chinese contents have been developed on behalf of Chinese Internet users, DNS is the "last-mile" that prevent Chinese netizens from using the Internet comfortably and conveniently.

Just as DNS, newly emerged IDNs might also cause problems in respect of intellectual property protection. However, non-English domain name disputes could be prevented and resolved through proper systems. Considering the business opportunity offered by IDNs in non-English speaking countries and regions, IDNs should not be regarded as harmful to IP owners.

Amengual

Value of IDNs: people in non-English countries may want to use internationalized characters for their domains.

Who will benefit: Domain registries and cybersquatters will definitely benefit from IDNs. Benefits for others do exist, but are too easy to be ruined: I can't speak for people in countries that naturally use multibyte characters, but in the ISO8859 world, people is already working with ASCII domain names and adding IDNs will only add costs and confusion if not done carefully.

Proof of benefits: Domain registries appear to charge per each domain, and IDNs do add a bunch of new possible domain names, so the benefit is clear. So much for the cybersquatters, as they can get IDN versions of domains that already exist.

Who will IDNs harm: people in non-ASCII countries operating sites with (perfectly understood and valid) ASCII-ized versions of the domains, if they miss their IDN equivalents. This case will also harm regular users who will never know if someone tells them to visit the "A" site or the "À", or the "Á". It is a problem for people who are able to correctly write the words, and will be even more of a problem for those who aren't.

Brunner-Williams

The value of IDNs is multi-part: it restarts an industry agenda that has been stalled since the major operating systems completed the "7-bit to 8-bit clean to multi-byte capable" program of the mid-90's, it extends the utility of text identifiers for numeric addresses to users lacking access to English language education, or for whom better identifiers exist, but require the use of characters not in the hyphen-plus-alphanumerics (aka "LDH" for "letters, digits, and hyphen") character repetoire, e.g., diacriticals present in both Spanish and French, or Asian characters.

The beneficiaries from the program of extending DNS identifiers are the DNS and infrastructure industries, which carry out a improvement in the service they deliver, and extend their markets; the DNS infrastructure users, both the "producers" and "consumers" of information, goods and services; and the concerned regulatory oversight bodies, for whom the problem space changes from a simple "business English" model to a pluralistic, multi-jurisdictional model.

Empirical proof of these benefits are available from operating system vendors who lowered their internationalization and localization (i18n and l10n, resp.) costs and extended their markets outside of the English language education markets in the early 1990's, from market data on the expansion of i18n/l10n capable systems outside of the English language education markets over the same period of time, and from studies on the transformation of institutions from simple contexts to more complex contexts.

Those harmed by extensions to identifiers used in the DNS are those who have "bet their business on English-Only", which includes some political groups and business entities in the United States and other English-using countries, and those who have a business model which cannot be modified to accommodate minor, incremental changes in infrastructure, regardless of whether the change is in the DNS, or in "internationalization", both, or neither.

CNNIC

It has been noted that the number of non-English speakers is increasing very quickly in the whole Internet Community. However, DNS, which was basically designed for English speakers, is significantly unfriendly to the non-English speakers. Since the Internet is the medium for communication, it should never be used as a test on people's English proficiency.

In China, over 80% of the population cannot read English (why should every body learn English?). Even though extensive Chinese contents have been developed on behalf of Chinese Internet users, DNS is the "last-mile" that prevent Chinese netizens from using the Internet comfortably and conveniently.

Just as DNS, newly emerged IDNs might also cause problems in respect of intellectual property protection. However, non-English domain name disputes could be prevented and resolved through proper systems. Considering the business opportunity offered by IDNs in non-English speaking countries and regions, IDNs should not be regarded as harmful to IP owners.

i-DNS.net

IDNs benefit both English and non-English using business and individuals. They help the former devise focused marketing strategies towards their non-English using target markets. By reaching out in a familiar language, business will find it easier to sell their products and services.

IDNs help non-English businesses reach out to their own language-speaking markets by enabling the registration and use of familiar company names, products, brands and marks.

It is undeniable that speaking the language of your consumers helps sell your products. We've already evidenced this offline, and do not expect a contradiction online.

As for non-English using individuals, we look to IDN as being an empowerment tool. By not having to struggle with an alien language to access websites, we believe that this will be the first step in getting them online and interacting on their own terms, in their own language.

The benefit IDN to e-commerce and the New Economy will be explosive - especially in countries where English is not commonly spoken or used at all.

Status Report: As above. Balance with more opportunity for cyber squatting.

Yabe The IDNs is useful and is easy to understand even for non-English speaking country people in the world. Especially, some non-European language speaking people should have benefits to utilize the domain names in their own languages.It is apparent that so many Chinese, Japanese and Korean people have been interested in IDNs.

2. Does the translation or transliteration of a trademark or other name constitute a violation? Does the answer to this question vary depending on the legal system? Do trademark treaties and other international agreements speak to this issue?

AIPLA In the AIPLA's view, the translation or transliteration of a trademark or other name capable of trademark protection could constitute trademark infringement and/or dilution, depending upon the degree to which a national legal system recognizes infringements and dilutive acts by a mark, name or domain name that is a foreign equivalent.
WALID With respect to questions 2 through 6 immediately above, WALID fully supports the work currently underway by ICANN and the World Intellectual Property Organization (WIPO) to address the legal and policy questions, including intellectual property rights, surrounding the domain name industry as a whole. We believe that the principles thus far developed have facilitated a more responsible approach in the DNS. The Uniform Domain Name Dispute Resolution Policy is an example of a mechanism that has increased user and industry confidence in the DNS. We would expect that with possible extensions, these principles should be applied to the IDNs space as well. WALID would look forward to participating in the on-going discussions of these issues.
Verisign IDN should not be an excuse to permit infringements. Since the circumstances will vary widely, however, the question whether the translation or transliteration of a particular trademark or other name constitutes a violation is one that is currently best addressed through the ICANN-approved Uniform Dispute Resolution Policy (UDRP). For example, an internationally famous mark may be treated differently than a locally recognized mark. In its introduction of IDN technology, VeriSign GRS made clear its commitment to fostering continued use of the UDRP by registrars, end-users, and third parties.
IPC

Trademarks can of course be registered in non-Roman script. The unauthorized use of the trademark as a domain name may amount to trademark infringement in the same way as unauthorized use of a Roman-based character trademark.

The transliteration or translation of a Roman-script trademark into a non-Roman-script language can amount to infringement. The question of trademark infringement varies according to the national law involved, but in very general terms is likely to depend on how the Roman-script trademark is known or referred to in the local language and whether the use of the translation or transliteration is likely to cause the public to believe that the trademark owner is associated with the word.

There are no trademark treaties or other international agreements which focus on this issue.

Neteka Neteka is not an expertise in intellectual property laws of the world, however we do believe that domain names should not be excluded from enforcement of the appropriate trademark laws.
Register.com As with other disputes involving the rights to a domain name, register.com would defer questions of trademark violations or other illegal acts to the appropriate entities in the legal and Intellectual Property communities and handle them in a way comparable to our current use of the UDRP.
Peacenet

2-1. Does the translation or transliteration of a trademark or othername constitute a violation?

Not necessarily. Whether the right of trademark is to be infringed or notdepends on the confusion of consumers. And confusion is decided depending on the similarity of trademark. Therefore, it depends on the similarity between the translation or transliteration of a trademark and trademark itself.

2-2. Does the answer to this question vary depending on the legal system?

The similarity could not be rigidly defined by legal system but to some extent depends on the level of consumers' understanding of the applied foreign language or a certain of transaction environment. So far as the institution of trademark is based on the framework of maintaining the competitive business environment by blocking out the confusion of consumers, this may be the same in most countries. In some legal system that adopts dilution provisions, individual rights of trademark owners is weighted rather than the confusion of consumers. But the dilution provisions could not change judging of similarity.

The similarity of the translation or the transliteration of trademark could be discerned by the generally applied principle of "sound, sight and meaning trilogy". This principle applies in Korea as well as the U.S., Japan and most European countries.

2-3. About UDRP

UDRP should be reviewed and complemented because it had never considered IDN. To judge the infringement of trademark in IDN particularly in the case of translation or transliteration of trademark, those who have the bilingual or multilingual capacity should participate in panels.Therefore, in future, there should be some measures for making service providers in geographic locations depending on language family or some obligatory requirements to have those qualified panelists in service providers of domain name dispute.

JPNIC There may be a violation in translation and transliteration depending on the legal system. For example, foreign trademarks are sometimes translated or transliterated into Japanese and the corresponding Japanese string may collide with other trademarks or trade names.
TWNIC

Since trademark right has the characteristic of territoriality, the answers to the above questions depend on the legal systems under which the trademark rights claim. Art. 6bis of Paris Convention, which is the most widely accepted international convention on industrial property, requests the member states to prohibit the use of "… a translation, liable to create confusion, of a mark considered by the competent authority of the country of registration or use to be well-known in that country as being already the mark of the person entitled to the benefits of this Convention and used for identical or similar goods". Art. 16 of WTO Trips Agreement extends the protection of well-known marks to "the goods or services which are not similar to those in respect of which a trademark is registered, provided that use of that trademark in relation to those goods or services would indicate a connection between those goods or services and the owner of the registered trademark and provided that the interests of the owner of the registered trademark are likely to be damaged by such use".

Under CDNC members' (e.g., CNNIC, TWNIC) Trademark Law and Implementing Rules of Trademark Law, use of translation of any well-known trademark of another party, liable to create confusion, is a violation of trademark right.

Amengual Transliterations of trademarks and other names will be constitutive of a violation in most cases. Most character sets are closely related to ASCII, and in most cases the extended characters are simply different versions of ASCII ones, just tailored to help in pronouncing them. But are the same letters, to all effects, as their ASCII equivalents.
Brunner-Williams

Translation or transliteration is inexact, technically these are many-to-many transformations of strings. Identifiers in the DNS have unique associations, hence a non-unique mapping transforms identifiers to non-identifiers.
The question posed appears to substitute "names" from the legal doctrine of marks, for identifiers.

However, there is a technical possibility which is germane to Question 2. See the technical discussion to Survey A for undefined terms used here. This is a possibility unique to the encapsulation transformation (ACE proposals), and may be ignored otherwise.

Assume that "Internationalized-Domain-Names" is a trademark, and assume that the encapsulation transformation (ACE mechanism) is simple rotation (rot13), (this avoids the necessity of using some non-ASCII characters and any known proposed ACE algorithm, and is as "ascii-gibberish" as any string with even one non-ASCII is after a real ACE algorithm is applied). The existing hyphen, digits, alpha repetoire:

-0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz
+---------+---------+---------+---------+---------+---------+--

is transformed to:

CDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz-0123456789AB
+---------+---------+---------+---------+---------+---------+--

hence:

Internationalized-Domain-Names

is transformed to:

V-5r3-n5v0-nyvBrqCQ0znv-Canzr4

The transformation and prefix ("13--") attachment of our example pseudo-IDN yields:

13--V-5r3-n5v0-nyvBrqCQ0znv-Canzr4

If the prefix is absent, so the string is no longer tagged as an ACE transform of some IDN, is the resulting substring within the scope of the general DRP?

What is the relationship between

These

www.Internationalized-Domain-Names.com
www.13--V-5r3-n5v0-nyvBrqCQ0znv-Canzr4.com

and these

www.Internationalized-Domain-Names-SUCKS.com
www.V-5r3-n5v0-nyvBrqCQ0znv-Canzr4-SUCKS.com

The non-technical reader should understand that users will "see" IDNs in several contexts, and will encounter in email, netnews, browser histories, etc, names of the form "www.tag--ascii-gibberish.TLD", and eventually will use prefix-free variations of reserved ACE'ed identifiers to construct alternate identifiers that plainly refer to the original ACE'ed identifiers.

The above response shows:

(a) that transliterations and transforms, if not one-to-one mappings, are outside of the DNS as surely as alternate roots are outside of the ICANN root. [Give the problem to John, he sings the fuzzy pontification ragga.]

(b) that ACE transformations yeild identifiers of the form tag--asciigibberish, and if the "tag--" prefix is removed, and the original string a "mark", the untagged transform, as well as the tagged transform, are probably "marks" as well. [I'm going to give this one to Jeff, he should enjoy it. It doubles the sale of ACE'd IDN "marks".]

(c) that "tag--asciigibberish" will leak, and become a lingua franca in their own right, forcing (b), above.

Now on to our last stop, (d).

However the translation or transliteration question is answered, the resulting answer is context-, as well as jurisdictionally-dependent.

Assume that the following set of octets (Unicode "U+hex16" notation) forms a string:

U+007e, U+0043, U+0061, U+0073, U+0068, U+005c

When viewed as ASCII encoding, the string is "~cash\"
When viewed as JIS X 0208:1997 encoding, the string is "(upper-hyphen)cash(Yen)".

[ASCII and Shift JIS "collide" on two code points, 0x7e and 0x5c.]

More generally, any set of encoded octets, in particular a set encoded in UTF-8, is not a string of characters until a context, usually in the form of a local code page, is provided.

The octets Öйú in charset=gb2312 are the characters for China (Chung guo). In charset=ascii these octets are undefined. In charsets=iso8859-{1-9} they are also undefined. In charset=big5 they are defined, but not as "Chung guo".

The interesting nuance that lack of local context provides is that as binary octets, even as UTF-8 encoded (or UTF-16) encoded octets, the putative "mark" cannot be constructed from the encoded format. If name server providers store "names" (identifiers!) in wire-code format (UTF-*), then they appear to have a rebuttable presumption to being incapable of determining what the character representation of the identifier is. This may reduce or eliminate the legal liability of infrastructure providers, particularly those not sharing the same local encoding as the hypothetically trademark infringing registrant.

Thus, jurisdictions will have to supply local context(s), and in general will not be able to supply global context, making the "jurisdictional shopping" exercise one of first "applicable local encoding" shopping.

[My thanks to Bill Manning who restated my thoughts related to item [30], now deleted (scope rules), from draft-ietf-idn-requirements-??.txt, and identified diminished infrastructure provider liability as a characteristic for ACE and UTF-* distinguishment.]

Trademark treaties and other international agreements do not speak to issues (a), (b), (c), or (d), discussed above.

CNNIC

Since trademark right has the characteristic of territoriality, the answers to the above questions depend on the legal systems under which the trademark rights claim. Art. 6bis of Paris Convention, which is the most widely accepted international convention on industrial property, requests the member states to prohibit the use of "… a translation, liable to create confusion, of a mark considered by the competent authority of the country of registration or use to be well-known in that country as being already the mark of the person entitled to the benefits of this Convention and used for identical or similar goods". Art. 16 of WTO Trips Agreement extends the protection of well-known marks to "the goods or services which are not similar to those in respect of which a trademark is registered, provided that use of that trademark in relation to those goods or services would indicate a connection between those goods or services and the owner of the registered trademark and provided that the interests of the owner of the registered trademark are likely to be damaged by such use".

China is a member state of Paris Convention. Under Chinese Trademark Law and Implementing Rules of Trademark Law, use of translation of any well-known trademark of another party, liable to create confusion, is a violation of trademark right.

i-DNS.net The registration of a domain name entitles the registrant to the use of the domain name in accordance with terms and conditions. Most policies that we are familiar with take the view that Registration does not confer any legal ownership or intellectual property rights over the domain name. The existence or extent of such legal rights depends on the intellectual property rights (e.g.. trademark rights) there may be over a particular name.

i-DNS.net endorses the World Intellectual Property Organisation's ("WIPO") Final Report of the WIPO Internet Domain Name Process (dated 30 April 1999) - a copy of which can be found at

http://ecommerce.wipo.int/domains/process/eng/final_report.html.

i-DNS.net's dispute policy at http://www.i-DNS.net/dispute.html is drafted & modeled along WIPO's recommendations and ICANN's Uniform Dispute Resolution Policy (UDRP).

In the event of a domain name dispute, we follow the processes on our dispute policy <www.i-DNS.net/dispute.html) and forward the matter to the relevant body for dispute resolution.

Status Report: As above. National variations in trademark law - translation/transliteration of a mark may infringe.

Yabe Even translation in certain non-European language might cause a trademark infringement in Japanese language in Japan. Especially, our Japanese have a tradition to describe foreign words by indicating the similar sounds to the original pronunciation in Katakana letters. The Katakana letters are commonly used as phonetic letters to indicate such similar sounds of original non-Japanese language. Therefore, even between just alphabet letters of certain English words and Katakana letters following its pronunciation in Japanese, a confusion and an infringement can be caused with trademark in Japanese.

In addition, the similarity of concept may have a chance of a trademark infringement according to the Japanese law and precedents. Therefore, we need to pay attention to possible similarity of IDNs with certain Japanese translation of foreign language.

However, the most serious problem is that some languages share common characters and letters in a certain area culture. For example, even though we cannot verbally communicate each other, Chinese, Japanese and Korean share same/similar characters and letters in our East Asian culture. I think that the same kind of thing can be found among Arabic speaking countries and African language speaking countries.

In this situation, even if the meaning would be different in each language but would use exactly same or similar, the conflict of IDNs registration and the chance of cybersquatting must be increased even between "different" language and "different" countries. The situation of cybersquatting may become much worse than now.

3. Will the existence of IDNs increase the incidence of cybersquatting? In what manner?

AIPLA The AIPLA believes that the existence of IDNs will increase the incidence of cybersquatting. Abusive domain name registration practices can occur with respect to the foreign equivalent of a well known mark normally appearing in Roman characters. See, e.g., Sankyo Co. Ltd. v. Zhu Jiajun, Case No. D2000-1791 (WIPO March 23, 2000). In Sankyo, the WIPO arbitration panel ordered the transfer of a two-character Japanese domain name which corresponded to Sankyo.com, based upon a finding that the domain name in question was registered and has been used in bad faith.
WALID With respect to questions 2 through 6 immediately above, WALID fully supports the work currently underway by ICANN and the World Intellectual Property Organization (WIPO) to address the legal and policy questions, including intellectual property rights, surrounding the domain name industry as a whole. We believe that the principles thus far developed have facilitated a more responsible approach in the DNS. The Uniform Domain Name Dispute Resolution Policy is an example of a mechanism that has increased user and industry confidence in the DNS. We would expect that with possible extensions, these principles should be applied to the IDNs space as well. WALID would look forward to participating in the on-going discussions of these issues.
Verisign While the overwhelming impact of IDNs on the Internet community will be favorable, the opportunity to register domain names in scripts other than the Latin alphabet, as well as the opportunity to register domain names in new top-level domains, can increase the incidence of cybersquatting. In other words, new and expanded opportunities to use the DNS can also lead to the possibility of increased illegitimate use. As with the current DNS, efforts must be undertaken to discourage and remedy such cybersquatting.
IPC

IDNs will increase the opportunity for cybersquatting. There are many hundreds of thousands of trademarks registered in non-Roman script languages and no doubt many more trademarks which are used without being registered. The ability to register a domain name in these languages will not only enable the trademark owners to register their own trademarks as domain names, but will also give the opportunity to cybersquatters to register the domain names. In addition, there will be the opportunity to register the non-Roman script equivalent (whether in the form of a transliteration or translation) of Roman-script trademarks.

Cybersquatting of IDNs is already an issue and several cases have been brought under the UDRP.

Neteka Cybersquatting is a necessary evil for the domain name system to continue to grow. The introduction of multilingual characters into the DNS system simply expands the possible namespace and therefore opens up more room for cybersquatting activities. In other words, simply by introducing more languages, there are more names for cybersquatters to squat.
Register.com At this time, we have not seen strong indications that IDNs would be more prone to cybersquatting than other gTLD domain names. While adding new languages may pose a linguistic challenge to trademark owners who are monitoring violations against their trademarks, it would be unreasonable to base stricter protections for IDNs than for other domain names on this concern alone. Likewise, we would hope that such concerns would not significantly slow down the implementation process of IDNs.
Peacenet

Basically, the introduction of IDN may have nothing to do with the increase of cybersquatting. Under the present condition that permit only limited subset of Latin characters, many alterations by translation / transliteration / hyphenating of the original name have been used.

However, if IDN is used, there is only one correct naming and only one domain name in its own language, therefore, the dispute could be rather vehemently increased.

JPNIC Cybersquatting will increase. This is because some Kanji characters are very similar to each other and most trademarks and trade names are registered in Japanese characters.
TWNIC Whenever new registration opportunity emerged, there could always be a rush of cybersquatting. Thus, either new gTLDs or IDNs will increase the incidence of cybersquatting. In the case of IDNs, cybersquatters would register the translations or transliterations of others' well-known marks, which may have been registered in English DNS.
Amengual

Definitely, yes. Again, I can't talk for the multi-byte world, but in character sets close to ASCII, it could happen that someone is operating the ASCII-ized domain name "A" and someone take the IDN "Á" domain. For people, "A" and "Á", are absolutely the same letter, but the DNS would send them to completely different sites. The IDN testbed already working has demonstrated that this case is unfortunately the rule and not the exception (just browse the NSI WHOIS database for some well-known Spanish domains).

What's more, if you can't have a trademark for your domain (which is the case for all domain names considered "generic"), you cannot invoke current UDRP to protect your domains.

Brunner-Williams Any increase in the available namespace(s) for which cybersquatting is defined will increase the incidence of cybersquatting.

Consider the following example from the proprietary Neteka extension to the alphanumerics-hyphen (aka "LDH" for "letters, digits, and hyphen") character repetoire. The ASCII character "$" is added.

Does the Jonny-Ca$h-Records-and-Tapes-Inc.com identifier cybersquat on the no-dollar-sign Jonny-Cash-Records-and-Tapes-Inc.com identifier? Yes.

The issue arises due to the number of similar characters in the repetoires generally considered in the context of internationalization of the DNS, not only is "DOLLAR SIGN" visually similar to "ESS" (and frequently used in lieu of "S" in non-DNS contexts), but there are hundreds, if not tens of thousands of characters in Unicode which are equally dubious.

Fundamentally the existence of IDNs increase the incidence of cybersquatting in three distinct forms, by increasing the total available namespace in which cybersquatters operate, by making a larger repetoire of characters available to form "off-by-one" ("invisibly different") identifiers and by making "alternate sense or form" identifiers possible in a vastly enlarged set of alternate character repetoires (languages).

CNNIC Whenever new registration opportunity emerged, there could always be a rush of cybersquatting. Thus, either new gTLDs or IDNs will increase the incidence of cybersquatting. In the case of IDNs, cybersquatters would register the translations or transliterations of others' well-known marks, which may have been registered in English DNS.
i-DNS.net Yes, when any new property is opened, there is speculation. However, this is not limited to the domain name system and happens in other walks of life.

We need to face this fact in a mature manner, and not let it be a barrier to change. Most registries have policies in place to deal with such matters, such that there is a line of recourse for all rightful name holders.

Status Report: As above

Yabe I believe that there is no doubt about the significant increase of cybersquatting after VeriSign started IDNs registration in last year. We can see the representative examples like Sankyo.com and Mainichi Newspaper at WIPO Arbitration Center. Some more Chinese domain name cases are also filed at WIPO's Center.

In Japan, a kind of "panic" was happened in last November. Because the major Japanese companies, which were diligent to register Japanese IDNs under VeriSign's services, found the truth that so many cybersquatting were already happened on their important and well-known corporate names, trademarks, trade names, etc. Without any attention to the trademark owners and legitimate owners of corporate/trade names in VeriSign's registration process for IDNs, the very disorganized situation was easily occurred.

Some registrant, a doubtful one for cybersquatting, tried to place their registration of Japanese.com/.net on the net auction and/or approached a certain company to buy it out with significant consideration. This is the truth happened in Japan We know that the similar chaos was found due to cybersquatting in US around 1994 and 1995. No empirical knowledge was utilized to avoid this virus-like spread of cybersquatting in IDNs.

4. What measures can be taken to minimize cybersquatting? Which of the following measures is most important - a "sunrise" period for pre-registration; a functioning WHOIS database; or a functioning UDRP system?

AIPLA The AIPLA views of equal importance a "sunrise" period for pre-registration; a functioning and robust WHOIS database; and a functioning and well-run UDRP system to minimize instances of cybersquatting.
WALID With respect to questions 2 through 6 immediately above, WALID fully supports the work currently underway by ICANN and the World Intellectual Property Organization (WIPO) to address the legal and policy questions, including intellectual property rights, surrounding the domain name industry as a whole. We believe that the principles thus far developed have facilitated a more responsible approach in the DNS. The Uniform Domain Name Dispute Resolution Policy is an example of a mechanism that has increased user and industry confidence in the DNS. We would expect that with possible extensions, these principles should be applied to the IDNs space as well. WALID would look forward to participating in the on-going discussions of these issues.
Verisign Measures to minimize cybersquatting include fostering the use of the UDRP, as described above. It also includes urging registrars during the testbed to consider deleting IDN second level domain name registrations upon receipt of a formal, written objection to the registration by any legitimate authority, including without limitation a trademark owner.
IPC

All three are important protection mechanisms.

A functioning and fully searchable WHOIS database is essential for dealing with cybersquatting, although language and technical issues need to be addressed.

Also, it is vital that IDNs should be subject to the UDRP. Dealing with disputes involving IDNs involves additional complications and the ability of the existing dispute providers to administer such disputes should be carefully analyzed and any amendments to the existing process to better accommodate IDN disputes should be considered.

The existing incompatible registration systems for IDNs have not included a sunrise for registered trademarks. We would favour a sunrise system for registered trademarks, but further consideration needs to be given to this issue and to the question of how existing registrations should be incorporated, if at all, into such a system.

Neteka The provision of a functioning and extensible system for multilingual domain registration and resolution is most important. Most companies are aware of the legacy cybersquatting issues by now and if multilingual names do work, it is likely that they will register for those that are relevant to their business. A backward compatible and functioning multilingual DNS plus the enforcement of the UDRP are the keys.
Register.com In our experience, the most effective tool against cybersquatting has been the UDRP. This has allowed trademark owners to handle trademark violations in a time and cost effective manner. We expect that the UDRP will be no less successful for IDNs than for traditional domain names. The UDRP is especially effective when coupled with a "sunrise" period that helps control cybersquatting by allowing preventative registration of domain names by trademark holders. In all cases, we feel that the introduction of new languages will not limit the effectiveness of these protective measures and that these measures should be similar to those used for domain names in Latin scripts.
Peacenet There might be no effective means to block out cybersquatting. Sunrise is not desirable because it limits the other users' rights. UDRP (or any *DRP which is currently being used for Latin character domain names by ccTLDs) is the only one possible option at this stage even in IDN.
JPNIC UDRP is essential. Additionally, sunrise period will facilitate the smooth start of new domain name space.
TWNIC

Both domain name registration agreements and DRP systems are effective measures to combat cybersquatting. Registration agreements provide a contractual-based control over cybersquatting. Contact details of registrants and open Whois database have been proved effective. Also, DRP offers a quick, cheap and almost fair channel for resolving domain name disputes.

Among the three measures--a "sunrise" period for pre-registration; a functioning WHOIS database; or a functioning UDRP system, it is hard to say which one is the most important, but a "sunrise" period for pre-registration may be caused more problems than those it prevents. The most dangerous of all, registries or registrars could get involved in domain name disputes caused by sunrise measure.

Amengual The "sunrise" period is important, but only if previously-existing domain names are considered at the same level as trademarks (for generic domains that simply cannot have the trademarks). The same for a Multilingual UDRP.
Brunner-Williams The measure that can be taken to minimize cybersquatting based upon exploit of visibly similar characters existing in Unicode is to put in place a new mechanism ("nameprep") which prevents visually similar characters from being used in identifiers. Returning to the ASCII "s" vs "$" case of the previous question, the name preparation process when applied to the string "ca$h" would return either an error (disallowed character) or "promote" the character to a matching character (map "$" to "s"), possibly with an end-user available warning of the character promotion.

A measure that can be explored to determine if it may offer a mechanism to manage cybersquatting based upon "alternate sense or form" is the restricted vocabulary approaches to automated translation, which has a literature and a product base in the localization market.

Of the measures cited in the question: "sunrise", whois, or a DRP, the first two relate to the technical operation of registries. Only the DRP has the possibility of having scoping properties that correlate to languages, rather than to namespaces.

Of the remaining two, whois is both a poor mechanism to provide sponsored or privileged access to registry data, and risks placing ICANN in a role that is antithetical to the Data Protection General and Special Directives of the European Union, and to the weaker counter-parts in the OEDC (Canada, Japan, etc.) and even the weakest FTC (US) sets of jurisdictions concerning privacy. This question could be better understood as:

an exhaustive database of registrant contact data, accessed by a means other than WHOIS

Finally, "sunrise" is in my opinion a poorer mechanism to manage bad-faith and speculative behavior than is an IP Claim Service.

CNNIC

Both domain name registration agreements and DRP systems are effective measures to combat cybersquatting. Registration agreements provide a contractual-based control over cybersquatting. Contact details of registrants and open Whois database have been proved effective. Also, DRP offers a quick, cheap and almost fair channel for resolving domain name disputes.

Among the three measures--a "sunrise" period for pre-registration; a functioning WHOIS database; or a functioning UDRP system, it is hard to say which one is the most important, but a "sunrise" period for pre-registration may be the most unimportant. The problems caused by a "sunrise" period are more than those it prevents. The most dangerous of all, registries or registrars could get involved in domain name disputes caused by sunrise measure.

i-DNS.net We do not believe that IDN's require any additional approaches that other new TLD properties require. We do not believe that there is a consensus in favour of sunrise periods.

There are already "technical aids" to this issue - 1) publicly available Whois database and 2) a functioning UDRP system.

Status Report: As above.

Yabe In order to minimize meaningless cybersquatting, all possible efforts---a well functioning WHOIS database, a "surise" period for pre-registration, a functioning UDRP system, has to be prepared.

If VeriSign properly paid their attention to a "sunrise" period for their registration of IDNs in last year, even under the test-bed, the multiplication of cybersquatting in IDNs could be prevented.

Unfortunately, there is no viable WHOIS data base for identifying each IDNs registrant in my understanding as of today. It is necessary to have a certain translator software to check which IDN is subject to whom. Even if we take into account some privacy issues of registrant, this kind of very inconvenient situation
cannot be justified for non-English speaking language trademark owners. Accepting IDNs, any registry/registrar should be serious and diligent to establish at least a "practically useful" WHOIS system.

UDRP also does not work well "quickly" for IDNs. It took many months more than originally designated to reach the panelist decision at Sankyo.com and Mainich Newspaper cases at WIPO Arbitration Center. In my understanding, there are not still enough number of panelists candidates retained by each Dispute Service Provider. The Providers are also frustrated by big needs for translation between English and other languages to handles IDNs cybersquatting. Even though a complaint, a registrant and a registrar are all located in Japan and the registration agreement is written in Japanese [the language to be used for UDRP should be Japanese under the current UDRP in this situation], WIPO Arbitration Center could not accept all correspondence only in Japanese and had still only several panelists from Japanese or other Asian character understanding people. As a result of that, the procedure took much time by comparing with a case for domain name just in alphabet. The situation might be also similar to other Dispute Service Providers.

5. What groups within and without ICANN should consider these policy issues? How should these groups proceed?

AIPLA The AIPLA believes that ICANN is proceeding in the right direction, namely, the establishment of a working group directed to identifying various internationalization efforts and the issues they raise, to engage in a dialogue with technical experts and other participants in these efforts, and to receive appropriate recommendations regarding IDNs. This working group should be comprised of members of the various DNSO constituencies and the Names Council. The working group should strive to develop a consensus of position and recommendations to the ICANN board, and issue a definitive report with concrete recommendations, as soon as possible.
WALID With respect to questions 2 through 6 immediately above, WALID fully supports the work currently underway by ICANN and the World Intellectual Property Organization (WIPO) to address the legal and policy questions, including intellectual property rights, surrounding the domain name industry as a whole. We believe that the principles thus far developed have facilitated a more responsible approach in the DNS. The Uniform Domain Name Dispute Resolution Policy is an example of a mechanism that has increased user and industry confidence in the DNS. We would expect that with possible extensions, these principles should be applied to the IDNs space as well. WALID would look forward to participating in the on-going discussions of these issues.
Verisign VeriSign supports the work underway by ICANN's Board of Directors, as well as by the Domain Name Supporting Organization (DNSO), to consider any policy issues raised by IDNs. VeriSign also supports the policy development efforts of organizations such as the Multilingual Internet Names Consortium (MINC).
IPC ICANN is proceeding in the right direction, namely, the establishment of a working group directed to identifying various internationalization efforts and the issues they raise, to engage in a dialogue with technical experts and other participants in these efforts, and to receive appropriate recommendations regarding IDNs. In particular, the views of experts in the field of intellectual property in the countries which will be most affected by the introduction of IDNs should be consulted. The working group should strive to develop a consensus of position and recommendations to the ICANN board, and issue a definitive report with concrete recommendations, as soon as possible.
Neteka ICANN should provide general guidelines on policy issues with the help of the members from the ccTLDs around the world. Unless there is an organization that would replace ICANN's authority for names and numbers, we do not see a need for these discussions to be outside of ICANN. ICANN should take the responsibility to identify and provide guidelines to resolve policy issues, much like the introduction of the UDRP.
Register.com Among the groups that should be involved in the discussion of these policy issues are the community members involved in privacy and IP concerns, GAC, and ICANN registries and registrars whose businesses are affected by these issues. The interested parties should work in cooperation with Internet standards setting bodies, such as the IETF, to ensure that technological standards are capable of addressing their policy concerns.
Peacenet ICANN should hear all voices coming from both within and outside its own community. To establish a working group in DNSO could be one best option for collecting diverse comments within ICANN. Also ICANN should contact with other groups outside of ICANN to be consulted. Furthermore, more cooperative relations should be formed with other organizations like MINC, CDNC and JET that have worked for IDN.
JPNIC We think ICANN should coordinate technical issues in parallel with technical standardization process through open and transparent discussion. This coordination, such as conducting registrars not to register ACSII strings with possible ACE headers, is essential to propel the IDN environment.
TWNIC There should establish a standing WG on IDNs policy issue in ICANN. The leader should someone who is a non-English speaker with significant legal and technological background and high profile in the non-English speaking Internet community. The working group should keep close communications with the Internet community.
Amengual Do not know about your internals.
Brunner-Williams The policy concerning character sets, equivalences, etc., may be considered by the IETF, which is better suited than any of the other PSO, and all of the ASO member entities. It may also be considered by a subset of the DNSO, the ccTLD Constituency and Governmental Advisory Committee have limited interest and experience in the area. This policy may also be considered by the Unicode Consortium, or by one or both of JTC1/SC2/WG2 (character sets) or JTC1/SC2/WG22 (internationalization).

The response to the questions of Survey A contains a technical discussion of character set issues.

The issue is both the ACE and UTF-8 ("client-side" and "server-side", resp.) approaches to IDNs rest upon a common character repetoire formally defined in ISO/IEC 10646-1:2000 International Standard -- Information technology -- Universal Multiple-Octet Coded Character Set (UCS), whose assignment of characters is synchronized with The Unicode Standard -- Version 3.0.

The Unicode Standard is a product of the Unicode Consortium, which is composed of printer vendors, and "glyph centric". This has a number of consequences, a few of which are overtly detrimental to the addition of IDNs to the DNS, and some are covertly so. It suffices for this policy survey to note that a private body dominated by printer vendors is institutionally disinclined to modify a standard that serves printer vendors well, regardless of how poorly such a standard serves the requirements for domain names.

There are 63 bytes available to construct multi-byte characters with, this is a limit which cannot be altered in the near term.

The ACE approach requires fewer than four (4) bytes to encode arbitrary characters in the UCS, yielding a maximum of 17-19 characters.

The UTF-8 approach requires exactly three (3) bytes to encode arbitrary characters in the USC, yielding a maximum of 21 characters.

It is possible to map all of the useful characters in the UCS repetoire to a much smaller space, and use only two (2) bytes to encode arbitrary characters in this mapped subset of the UCS, yielding a maximum of 31 characters.

The policy concerning "alternate sense or form" (translation), may be considered by the DNSO IP Constituency, or by invited experts from the localization industry if no corresponding organization can be found.

The policy concerning "sunrise" (and IP Claim Service) may be considered by the IP, Registrar, and the (cc|g)TLD Registry Constituencies of the DNSO.

The policy concerning "WHOIS" may be considered by the (pending) WHOIS WG of the IETF, in conjunction with the Registrar, the (cc|g)TLD Registry Constituencies of the DNSO, the five RIRs of the ASO, and the Governmental Advisory Committee.

The policy concerning "UDRP" may be considered by the (cc|g)TLD Constituencies and Governmental Advisory Committee, WIPO, and recognized language authorities.

The character set policy group should proceed by fully specifying characters which are not allowed in IDN identifiers (nameprep), as the IETF nameprep activity may be accurate but incomplete. It should then begin a program to optimize, or replace Unicode with a character centric alternate, which has the property of greater effective language coverage, blocks of code points that may be "delegated" to language authorities that are not yet included in the Unicode glyph set, and optimal for use in small strings such as the current 63 byte limit for names in the DNS.

The translation policy group should document existing trademark translation norms and solicit comment from the machine translation experts in the restricted language area of the localization industry.

The WHOIS policy group should review the ICANN WHOIS Survey data, the policy statements of concerned privacy and data protection parties, and contribute to the (pending) IETF WHOIS WG.

CNNIC There should establish a standing WG on IDNs policy issue in ICANN. The leader should someone who is a non-English speaker with significant legal and technological background and high profile in the non-English speaking Internet community. The working group should keep close communications with the Internet community.
i-DNS.net

The DNSO should be an interested party. However we do not believe that IDN's require special Policies to be developed at the ICANN level.

Status Report: As above, including DNSO, IETF, CDNC, JET etc.

Yabe

It is necessary for Intellectual Property specialists from certain language used for IDNs to have more discussion and cooperation with Internet engineers, registrars and registries. For example, regarding IDNs, the mentioned-above people from the same language roots should have more chances to express their thoughts and idea at ICANN activities.

Particularly, a kind of MINC's effort should be considered inside of ICANN.

6. What other legal and policy issues are raised by IDN? How should ICANN address them, if at all?

AIPLA The legal and policy issues raised by IDNs, as stated above, involve enhanced opportunities for cybersquatting by foreign equivalents of trademarks or service marks normally displayed in English and/or Roman characters. ICANN should address this issue by modifying the Uniform Dispute Resolution Policy ("UDRP") to specifically recognize bad faith registration and use of domain names by foreign equivalents, that is, domain names in other languages which, translated or transliterated, would correspond to trademarks and/or service marks that normally appear in English and/or Roman characters.
WALID With respect to questions 2 through 6 immediately above, WALID fully supports the work currently underway by ICANN and the World Intellectual Property Organization (WIPO) to address the legal and policy questions, including intellectual property rights, surrounding the domain name industry as a whole. We believe that the principles thus far developed have facilitated a more responsible approach in the DNS. The Uniform Domain Name Dispute Resolution Policy is an example of a mechanism that has increased user and industry confidence in the DNS. We would expect that with possible extensions, these principles should be applied to the IDNs space as well. WALID would look forward to participating in the on-going discussions of these issues.
Verisign This survey, highlights key policy issues relevant to IDN. We therefore hope that many of the other entities (commercial and ccTLDs) offering the opportunity to register IDNs will also respond.
IPC The IPC is limiting its comments to those affecting the rights of intellectual property owners.
Neteka ICANN should definitely provide domain registration policy guidelines for multilingual names. Professionals such as WIPO should be consulted for legal issues such as trademark or intellectual property issues.
Register.com

One policy issue raised by IDNs is implementation and communication between registries and registrars.in the IDN space. Ample and equal notice from registries to registrars regarding testbeds and other new developments is important in creating an even playing field onto which new technologies will be introduced. ICANN can help achieve better communication by actively encouraging dialogue between registries and registrars. If such voluntary measures do not succeed, ICANN may need to consider using its authority to ensure that individual actors do not harm the overall interests of the Internet community.

A second policy concern is the involvement of non-English speakers in the policy development and implementation processes. Representatives from communities using non-Roman languages should be thoroughly consulted as IDN decisions are made. This includes making documents, such as the current survey available in a variety of languages.

Peacenet

6-1. What other legal and policy issues are raised by IDN?

If there are some language families that could bring about some confusion in naming expression even among different languages, some attentions should be paid in legal, technical or other policy aspects.

Domain name is being used for web surfing as well as functioning as anidentifier for confirmation in most transactions including electronic commerce. Therefore, the security and the verifiability are also very important issues. Then, IDN could have many similarly looking characters even among different dozens of languages. Those could be used for fraud and consequently threaten the security.

Therefore, some control over registration permission or dispute resolutionis required so that this collision could be minimized. To solve this technically, if similarly looking characters are inserted in name space, the one unified representative character could be replaced as anormalization process. If this kind of policies could be adopted, it could mitigate the problems.

However, Japanese katakana ni, it looks the same one as Chinese character number two. So, it is not easy to determine whether those are to be same or not. Many cases like this will be found if studied.

6-2. How should ICANN address them, if at all?

(= 5)

JPNIC

Among others, the following issues are thought to be the most important:

1) Internationalization of TLDs. Although we need more discussion, we tend to state TLDs should not be internationalized because language and countries(cc's) do not represent the same set of users.

2) Alternative roots. Internet community must not have alternative roots. And ICANN should try to prevent from alternative roots.

3) Languages for UDRP process. The language, which is implied by the "domain name," should be able to be used in the UDRP process even if the domain name is being registered in a jurisdiction with English.

TWNIC

UDRP might not be fully applicable to IDNs. Thus, UDRP itself may need modulation to some extent. Further, the operation system of UDRP may also need reform. Current UDRP service providers (WIPO, NFC, CPR and e-Resolution) are undoubtably competent of dealing with English domain name disputes, but their capability and resources for dealing with IDN disputes are fairly suspicious. Should ICANN consider delegating dispute service providers in different language communities?


In my humble opinion as a besieged trademark holder the limited viability of the proposed extensions will prove, in the end, not to be worth the effort now being undertaken. The primary beneficiaries of these proposals are the speculators and cybersquatters.

Bird In my humble opinion as a besieged trademark holder the limited viability of the proposed extensions will prove, in the end, not to be worth the effort now being undertaken. The primary beneficiaries of these proposals are the speculators and cybersquatters.
Amengual Violations caused by transliterations are probably the only problem for character sets close to ASCII.
Brunner-Williams The legal framework WIPO, then ICANN, brought to the question of identifiers and marks has a wider scope than "business English". However, the addition of character repetoires in ISO 10646 extends the "linguistic space" available to identifiers beyond the WIPO framework. Identifiers in the internationalized DNS will use characters in Inuktitut, common to Greenland, Nunavut, Northern Canada, and Alaska, characters in Algonquin, common to Southern Canada and the North Eastern United States, characters in Dineh (Navajo), and characters in Tohono O'odham, common to Arizona and Northern Mexico.

These character repetoires have authoritative references and legal institutions which are not "states" in the WIPO nation-state (Treaty of Paris) framework.

To which jurisdiction, to what theory of dispute resolution, shall parties in conflict over identifiers using Dineh characters forming Dineh words avail themselves to? The issue is no longer which national court system or national intellectual property bar association "owns" French or Chinese. The issue is are there "languages" for IDNs for which the language authority and associated legal institutions are not directly available under the ICANN system?

The policy framework ARPA, then ICANN, brought to the question of identifiers and users has a wider scope than "unambiguous resolution". Again, expansion of the character repetoire extends the identifier "linguistic space" beyond the ARPA framework. Identifiers in the internationalized DNS will use words and phrases in ideomatic Spanish, French, Chinese, and hundreds of minor or endangered languages.

These words or phrases are parts of living languages who's user communities are not generally classed as "privileged early access" to the Internet, or omputers, or even to POTS. They are on the wrong side of the Digital Divide.

To what purpose, to what theory of utility, shall parties use words from the languages of minorities or late adopter majorities? The issue is no longer the "business use" of languages to form identifiers. The issues is are there "literacies" for IDNs for which the use will advance basic and advanced literacy, promote economic and cultural development, and advance the ICANN agenda?

ICANN should address the jurisdictional issue where languages and states do not coincide. It can address the issue by affording direct access to the dispute resolution process to jurisdictions and language authorities which do not have direct access to the International legal system.

ICANN should address the linguistic issue where business and literacy do not coincide. It can address the issue by adopting a policy of linguistic, as well as geographic diversity, and a policy of identifiers as tools for the delivery of literacy, as well as for the delivery of business.

CNNIC UDRP might not be fully applicable to IDNs. Thus, UDRP itself may need modulation to some extent. Further, the operation system of UDRP may also need reform. Current UDRP service providers (WIPO, NFC, CPR and e-Resolution) are undoubtably competent of dealing with English domain name disputes, but their capability and resources for dealing with IDN disputes are fairly suspicious. Should ICANN consider delegating dispute service providers in different language communities?
i-DNS.net

See 5.

Status Report: Overall communication needs to be improved. Modulation of the UDRP to handle IDNs.

Yabe

Currently, some major ccTLD registries/registrars are trying to be involved with IDNs for their own local domain names: .jp, .cn, etc. In this trends, a more complicated situation has been happened, for example "non-English.non-English" IDNs. I wonder whether this "very local language oriented" IDNs would become harmful for the Internet in future. I am afraid of extreme locality to split the Internet and to be more sympathized for the alternate rooting.

ICANN should be very careful to watch this situation and to properly respond the situation.

7. Presumably there will be a demand for top level IDNs, i.e., [IDN].[IDN]. What role should ICANN play with respect to the selection of these top level names and their registries? What should be the role of the country code registry of the country where a particular script is a native script? And what if a particular script is a native script in more than one country?

Brunner-Williams There is a demand for IDN.IDN, an example in point is .CN (People's Republic of China), which is ."chung guo" ("Middle Kingdom", the Chinese name for the historic and modern Chinese State).

Where a ccTLD operator has a bona fide request for addition of national language alternates to ISO 3166 two-letter country codes, such as the two Chinese characters for "middle" and "kingdom", ICANN should add the IDN form(s) to the DNS root.

Thus far IANA, and now ICANN, have only made use of the ISO 3166 two-letter country codes, not ISO 639 two- and three-letter language codes, nor any longer form, e.g., ".English", or ".Arabic".

ICANN should solicit proposals for top level domains which meet the needs of "endangered" or "minor" or "lesser taught" languages, or those that seek to provide a shared TLD infrastructure for such languages, such as a TLD for the indigenous languages of the Americas.

ICANN should encourage technical cooperation between registry operators who serve shared language communities, such as the Chinese Domain Name System (CDNS) Joint Engineering Team (JET), which coordinates between the .CN, the .TW, the .HK, and the .SG registry operators.

Chon Here are some of my thought on the internationalized domain names (IDN) and ICANN. ICANN should have started discussion on IDN earlier due to its urgent nature and relevance to the fundamental nature of ICANN with administration of TLDs and root servers.

1. Boundary of ICANN

Where is the boundary of ICANN's coverage? Are the ASCII TLD the only TLDs ICANN is concerned? Are the internationalized TLDs part of ICANN activities?

2. Internationalized TLDs

It would take much time and effort to come up with appropriate policies on the internationalized TLDs, and there are deployment of the internationalized TLDs now. Is ICANN late again like the new gTLD deployment?

There are the following and more issues, and serious study has to commence as soon as possible;

(A) Multilingual Root Server

In order to handle internationalized TLDs, the root servers have to be ready to accommodate them before the deployment takes place.

(B) Multi-country Language

Internationalized TLDs based on the language which is spoken in more than one country such as Chinese, Arabic, French, or Russian need additional policy to handle the multi-country issues.

(C) Non-country Language

There are languages such as Tamil and Kurdu which are not the major language of any country. Handling of such languages may require additional elaboration. ccTLDs may not be relevant in this case, and we may have to create "language code TLDs".

3. External Definition of Internationalized Domain Names
The internationalized domain names as well as the internationalized non-domain names should be defined in name space so that we can have proper perspective on the internationalized domain names and its name space.

4. Standards Related to Internationalized Domain Names

The interim report is focusing on the standardization only on the idn so far. Other standardization activities should be addressed such as email, whois database, web and cname.

5. Statistics

We need statistics on the current deployment. According to our record, there are around 2.1 million domain names registered. Case studies on some of ccTLDs and gTLDs would be needed, too.

6. Coordination

I don't see much coordination on the internationalized domain names within ICANN nor with external organizations. This is the topic that we need good coordination internally and externally.

Within ICANN: Board, Names Council, Constituencies, RSSAC, Protocol Council

External: IETF, MINC, WIPO, W3C, ISO, Unicode, ITU,...

7. Multilingual Dispute Resolution Policy

Internationalized domain names may need different dispute resolution policy from the ones for ASCII domain names, and this should be studied.

8. Internationalization of the Internet

The internationalization of domain names is part of the internationalization of the Internet, which is a major undertaking and would take long time. Some of the issues are of short term, and we need to address the long and short term issues properly.

CNNIC (Answer to B-7) Since the fundamental purpose of deployment of IDNs is to serve the demand for non-English speaking Internet communities, IDN management should not be fully controlled by commercial interests. Where ICANN making the policy of IDN management, those language users should be given the opportunity to voice their opinions on relevant IDNs. Existing ccTLD managers are in the vital position in respect of IDN management. Because those managers are delegated by ICANN and approved by relevant governments. They are in close relationship and effective communication with local Internet community where a particular language is a native language. Therefore, where IDNccTLDs are concerned, it is desirable to retain those managers' management status and system when deploying IDNccTLDs. Since many countries or regions may share the same language, and IDNgTLDs are needed. In order to create a forum for the users of a language to voice their opinion, the Internet community that shares the same language should be encouraged to form cooperative organization. Consensus on IDN gTLDs may be formed on the basis of these cooperative organizations. The solution based on consensus should consider the habit, custom and culture of specific language community. The ccTLD manager of the country or region in which majority of the users of a language reside may act as the coordinator of the cooperative organization of the specific language, but generally the organization should be open to all the users of that language.
CNNIC (Narrative Response) We noted the ICANN's announcement published on August 25, 2000. We agree with ICANN that it is very important that the Internet evolves to be more accessible to those who do not use English-language character sets.

We also noted the Status Report of ICANN IDN WG released on June 1, 2001. According to the Report, respondents to the Survey launched by ICANN IDN WG, generally had a very positive attitude towards IDNs. They observed that most of the world's population does not use Latin script as its native script. IDNs, accordingly, will increase access to and use of the Internet. Additionally, IDNs will increase access to this population by businesses and organizations that now are constrained by Latin script domain names.

We agree the principle of development of IDNs that the internationalization of the Internet's domain name system be accomplished through standards that are open, non-proprietary, and fully compatible with the Internet's existing end-to-end model and that preserve globally unique naming in a universally resolvable public name space. A fundamental requirement in this work is to ensure that the DNS continuously allows any system anywhere to resolve any domain name, without disturbing the current use and operation of DNS.

We, however, worry that without proper IDN management policy, there is the danger of "balkanization" of Internet. During the policy making process, we hope ICANN could enhance the harmonization with local Internet communities that are interested in and influenced by IDNs.

We are of belief that the following points should be highly noted in developing IDNs.

1. The development of MDN should not only ensure the stability and compatibility of the current Domain Name System, but also should guarantee the interests of those language users and respect the policy mechanisms of local society, such as politics, economy, legal system and culture.

2. IDNs are not simply the issue of Technique, but more relevant to Management. Since the fundamental purpose of deployment of IDNs is to serve the demand for non-English speaking Internet communities, IDN management should not be fully controlled by commercial interests. When making the policy of IDN management, those language users should be given the opportunity to voice their opinions on relevant IDNs.

3. Those cooperative organizations formed by the Internet communities who speak native languages other than English should be encouraged to play an important role in relevant IDN management, provided that those organizations operate within the current DNS and under the coordination of ICANN.

Considering the complexity of management of IDNs, we suggest to set easy before hard, gradually open and push forward IDNs by stages and levels.

STEP ONE: IDNs and ccTLDs

Significances:

  1. Intensify ICANN's authoritative status;
  2. It is easy for ccTLD managers to provide IDNccTLD registration services if ICANN add IDNccTLD into the root server.
  3. Easy for people to use IDNs. Cybersquattings will be reduced, for people no longer apply for IDNs from different registries.

    Level One: IDN.IDNccTLD

It's suggested that ICANN establishes a list of IDNccTLDs.

Definition: IDNccTLDs refer to TLDs in IDNs. Generally, IDNccTLDs should be the country names in the national languages corresponding to existing ccTLDs. For example, China's IDNccTLD should be [.China in Chinese].

Significance: IDNccTLDs are not only identifiers, but also relevant to national identities. ICANN should be very cautious when dealing with such issues. Feasibility study may be needed before any policy is made.

Registry: Existing ccTLDs managers are in the vital position in respect of IDNccTLDs. Those managers are delegated by ICANN and approved by relevant governments. They are in close relationship and effective communication with the governments. Where national identities are concerned, it is desirable to retain those managers' management status and system when deploying IDNccTLDs.

Characteristics:

  • IDNccTLDs should be completely on voluntary basis. Existing ccTLDs managers may apply for a IDNccTLD or choose not to do so.
  • Each ccTLD manager should only be permitted to apply for one IDNccTLD.
  • It is each ccTLD manager that decides, under the approval of relevant government, the name or language of the IDNccTLD applied. A country may have multiple names (or abbreviation) or may use multiple languages. Deciding the name or language used in the IDNccTLD should be the internal affairs of that country. It is suggested that ICANN should defer the delegation of such IDNccTLD until the relevent country has selected one.
  • ICANN may limit the length of the scripts of each IDNccTLD. There should be a procedure to address the issue of the duplication of applied IDNccTLDs.

    Level Two: IDN.ccTLD

Significance: we noted that ICANN recognizes the practices of IDN.ccTLD provided by ccTLDs managers .We acknowledge that IDN.ccTLD is an expedient solution for the demand of IDNs of local Internet community.

ccTLD Managers: It should be in ccTLD managers' discretion to decide whether to provide IDN.ccTLD services.

ICANN Coordination: In order to prevent disputes between such ccTLD managers, we suggest that ICANN coordinate such mixed IDN services and ensure that one ccTLD manager merely provide one local language IDN service under the ccTLD.

Common Ground for the Two Levels: DRP

Localization: Since IDNs under ccTLDs are in local languages and registered through ccTLD managers, DRP should be more localized, and managed by local institutions.

DRP Systems: Once the IDNs under ccTLD system are deployed, all the existing DRP systems (court proceedings, arbitrations or ADRs) in each ccTLD may extend to them to resolve the disputes caused by abusive registrations.

STEP TWO: IDNs and gTLDs

Level One: IDN.IDNgTLD

It's suggested that ICANN create IDNgTLDs after effective consultation with non-English speaking Internet communities that would like to use gTLDs in their own languages.

Definition: IDNgTLDs refer to gTLDs in IDNs. For example, corresponding to gTLD ".com", Chinese users may prefer [.com in Chinese], while Japanese users may prefer [.com in Japanese].

Significance: IDNccTLDs cannot fully satisfy the demand for IDNs, because many countries or regions may share the same language, and IDNgTLDs are needed in these language communities.

Current Problems: At the present time, several commercial interests around the world start to recommend new IDNgTLDs and provide corresponding registration services.This kind of IDNgTLDs developed by the different commercial interests, each does things in its own way and no uniform and normative DRP system.So the registration under these IDNgTLDs will inevitably result in a lot of cybersquattings and cause the perplexity in the field of existing Internet DNS.

Coordination: Since IDNs are developed to serve for non-English speaking Internet communities, creation of IDNgTLDs should not merely be controlled by commercial interests. Language users' interests must be taken into account.

Consensus: In order to create a bottom-up system for the users of a language to voice their opinion and input their suggestions, the Internet community that shares the same language should be encouraged to form cooperative organization. Consensus on IDNgTLDs may be formed on the basis of these cooperative organizations. The solution based on consensus should consider the habit, custom and culture of specific language community.

ccTLD Managers: The ccTLD manager of the country or region in which majority of the users of a language reside may act as the coordinator of the cooperative organization of the specific language, but generally the organization should be open to all the users of that language.

Registry: Each cooperative organization, after consultation with ICANN, may act as the nomination body for the registry of the IDNgTLD in specific language based on the consensus of the relevant community. ICANN may delegate the registry after receiving the nomination from relevant cooperative organization.

Level Two: IDN.gTLD

Current Situation: Some gTLD manager has provided IDN.gTLD services to the public. There exist serious abusive registrations and improper registration management in such services.

ICANN Management: We suggest, rather than taking "let it be" policy, ICANN should cooperate with non-English speaking cooperative organizations to regulate such gTLD manager's practices, especially in the aspects of registration policy and DRP issues.

Common Grounds for the Two Levels: DRP

Characteristics: The DRP should keep the balance between the characteristic of gTLDs and the characteristic of IDNs. Therefore, a unified DRP may not provide desirable solution for the disputes in different languages. The DRP must take into account the various elements of the culture, language and legal system relating to IDNs.

Solution: The DRP for those gTLDs may not be "localized", but should be "regionalized" based on language communities. Non-English speaking cooperative organizations should develop DRP for domain name disputes in their own languages, and delegate dispute service providers.

i-DNS.net

Yes, there is already such a demand and I-DNS>net is a solution provider for a number of ML TLD's. In the absence of ICANN, I-DNS.net works collaboratively amongst the key players in local communities to establish an appropriate choice of TLD string before launching services.

We'd envisage ICANN taking similar involvement, to hand-in-hand with MINC, AINC, JDNA (and other regional or global legitimate consortiums/bodies formulated to facilitate the development of Internationalized Domain Names) as well as in-country authorities and linguistic experts in the selection of these multilingual variants of TLD's. As per its credo, ICANN should work closely with all Internet stakeholders and take on the role of a coordinating body for all local policy-formulating organisations.

For registries, ICANN should objectively evaluate the performance (e.g. operational know-how and expertise, financial capability, technical and infrastructure capability, etc) of current IDN service providers and select one (using a list of objective criteria) that will be able to best serve the Internet community at large. Should all current providers fail the selection process, ICANN may open the TLD for biddings/tenders by potential Registries.

We envisage that ICANN's ML policy formulation involvement in country code Registries and in-country Internet authorities would be similar to that of the current ASCII TLD's; i.e. country code Registries determine policies for ccTLDs. Usage of country code top-level names should also be set and determined by the country code Registries and in-country Internet authority.

Where a particular script is a native script, in-country Internet authorities and linguistic experts should be consulted during the selection process on the usage of top-level names. If a particular script is a native script in more than one country, ICANN and the bodies concerned, should play a coordinating role to achieving general consensus between the countries' Internet authorities, linguistic experts and market forces on the usage of a particular generic top-level name.

TWNIC (2)

We suggest:

1. ICANN remain to play the role of single authoritative root server manager and to maintain the Internet stable operations.

2. ICANN should encourage the testbed for [IDN].[IDN] .

3. The current ccTLDs managers have the priority right to decide whether keep going to provide testbed service and to be the corresponding [IDN].ccTLD or [IDN].[IDNccTLD] testbed manager.
4. For [IDN].gTLD and [IDN].[IDNgTLD] testbed, we suggest the certain language speaking country should be considered in first priority as the manager.

5. If certain language is a native language in more than one country, ICANN should encourage they form a cooperative organization to make consensus. And we suggest ICANN should fully respect the culture and customs of the region where majority of certain language users are living in.

8. Several respondents indicated that cybersquatting should be fought through greater use of regional dispute resolution procedures. Should ICANN encourage the development of these RDRPs? Should it establish criteria for RDRP providers? Or should it encourage the existing UDRP providers to increase the number of panelists familiar with non-Western languages?

Brunner-Williams

Yes. ICANN should require existing UDRP providers to increase the number of panelists familiar with non-Western languages, and where possible, delegate to DRP regional providers.

CNNIC

ADR has been proved to be an efficient and effective mechanism for resolving domain name disputes. UDRP was created basically because of the global nature of gTLDs. In respect of ccTLDs, RDRPs, which was created and are operating under local legal systems, are more feasible. Although there may be or should be consensus to some extent on the framework or principles of all these ADR systems, a single DRP can never be satisfactory in any situation. In cases of IDNs, DRPs should show more flexibility and diversity. More localized DRPs may be more suitable to resolve the disputes of domain names in local languages.

i-DNS.net

We do not see deployment of IDN' creating any new class of disputes, and would not initially be supportive of any new approach to handling disputes for IDN's.

In our experience IDN's raise a couple of unique issues, over and above the current DRP process.

First, the UDRP is an ENGLISH based system, menu driven. Where name are registered in local languages, the Dispute Resolution providers need to adapt their systems to converse in the appropriate language, including handling the ML domain name.

Secondly, the choice of language and of documentation under which the Dispute is managed may change to the local language, rather than English. We understand that this practice occurs in the current process amongst certain providers. Use of IDN's may accelerate this process.

Thirdly, DRP's may need to increase their local language skills in order to converse in the local language. We envisage that some dispute providers may choose to specialize I particular languages. Conversely, some may withdraw from managing same languages and maybe establish behind the scenes arrangements with DRP's that handle that language. This may be a behind the scenes change, but could reduce the number of DRP's available.

We do not see the need at this point to extend UDRP into a regional framework. However, it may be that ICANN play a role in the development of general guidelines, a framework and criteria for RDRP providers. From this it may be that a group of "regional experts" arises, although we do not see a need to regulate for this.

TWNIC (2)

We suggest ICANN should respect ccTLD managers to establish their own appropriate RDRP. At the same time, we also encourage UDRP providers to collaborate with ccTLD RDRP providers to solve non-western languages dispute problem.

 


Comments concerning the layout, construction and functionality of this site
should be sent to webmaster@icann.org.

Page Updated 29-Aug-2001
(c) 2001  The Internet Corporation for Assigned Names and Numbers. All rights reserved.