Babel’s back

It’s a story that’s been doing the rounds for some days now. Internationalised Domain Names, or IDNs, have come home: it will soon be possible to own domain names in a handful of Indian languages, including Hindi, Tamil, Bengali, Punjabi and Sanskrit.

The National Internet Exchange of India (NIXI) announced the launch of IDNs in five to six languages in the first phase of an ambitious project beginning early next year.

(NIXI is a non-profit government agency that serves as an operational meeting point of Internet Service Providers (ISPs) in India. NIXI also operates the official .IN registry.)

Similar to the launch of the .IN registry, NIXI plans to open the IDNs first for trademark owners in a Sunrise Period starting January 2008. Registration for the general public is expected to start around March, 2008.

IDNs actually convert non-ASCII characters to ASCII (character encoding based on the English alphabet, used extensively in computing) in order to preserve compatibility with the existing infrastructure.

Even as Indian entrepreneurs start to celebrate, there are some obvious problems with a multilingual IDN system, which NIXI additional CEO Rajesh Aggarwal acknowledges in the ET article cited above:

“The same letter may exist both in Tamil and Hindi. Also, within a language it can be represented visually through different set of Unicode characters. We are trying to make the IDNs phishing and spoofing proof,” Aggarwal added.

Currently, any letter in a Devanagiri, Gurmukhi or any other regional script font can be created visually using different codes.

Simply put, if is created using one script, the same can be created visually by using a different set of codes. It may lead to a rise in phishing or spoofing of identities.

However, a policy is being prepared to handle such eventualities.

“To counter this, we will block alternate identities of the same name, which can be created by different codes, except the original identity for every user. We will have a scientific formula on how the visual representation of each character in the domain name will be made. Also, we will have a simple dispute resolution policy,” he [Aggarwal] adds.

In a related development, the Internet Corporation for Assigned Names and Numbers (ICANN) has started live testing of Internationalized Domain Names in 11 non-English languages, including Hindi and Tamil, which allows Internet users to key in the entire domain name in one language. (Currently, domain name suffixes, e.g . “.com” and “.in” are only available in Latin characters.)

This ICANN project has been in the works since 2000 and has taken its time to come to some palpable fruition. Some time ago, China set up its own system to allow its citizens to access net domains written in Chinese characters, fed up waiting for [ICANN] to approve an official way of using non-Roman alphabets in domain names.

So who’s celebrating? Surely, it’s the countries where English is not the primary language of communication. The obvious argument in favour of IDNs is the prospective expansion of the Internet to parts of the world that have been on the other side of the Germanic divide. If one can already have a site with content in Tamil, it has always seemed quite silly to have to key in a domain name with Latin characters.

At the same time, the logistical enormity of setting up a system of this kind is pretty overwhelming. At the very least, trademark owners will have a tough time registering multi-lingual identities.

From the regulator’s perspective: while NIXI might be able to deal with phishing and cyber-squatting in languages it is immediately overseeing, e.g. Hindi and Punjabi; it might get more complex in cases involving non-Indian languages such as simplified Chinese.

From the trademark owner’s perspective, a very simple but daunting question: how many languages does one keep tabs on? And how does one do this? For a reputed international trademark owner, this might be an easy task. But what of the smaller, emerging entrepreneurs who do not have the resources ( e.g., international offices, linguistic expertise) to pull this off?

There is also the issue of languages that share characters and scripts, e.g., Hindi and Sanskrit share the Devanagari script, which allows the same character to be represented by two different codes, welcoming hackers and their ilk with open arms. (See homograph spoofing). Although the existing defence mechanism lies in software design itself (e.g., web browsers do not support certain kinds of domain names), its acceptance as the best possible method of protection is still under question. And this debate will surely trudge on in the months to come.

Cyber-squatting, and related trademark dilution, also comes into play with regard to IDNs in languages that are used simultaneously in different countries. A case that stares in the face is that of Tamil, which has official language status in India, Singapore and Sri Lanka. The domain name-trademark relationship, although it exists in principle, is yet to be strengthened in practice. With different IP laws in different jurisdictions, this correlation is going to become weaker than before. And of course, one hasn’t even begun thinking of simpler administrative issues. (Hypothetically, if prior registration of a trademark is a criterion for domain name registration, there are significant time-lags in the certification and registration processes in different countries; with its related problems).

On an endnote, though, a new wave of regional Top-Level Domains has opened up for registration, with the “.asia” suffix up for grabs. It has taken its time, but to force a cliché, Asia is surely the flavour of the season. And more exciting still, although perhaps fanciful, we might yet live to see the lingua franca of the Internet change…


About The Author

Leave a Comment

Scroll to Top