Earlier this week, technical and legal experts from ICANN, Deloitte and IBM met in Brussels with their counterparts from registrars and registry service providers representing as much as 90% of prospective gTLD registries, to discuss the technical implementation of the Trademark Clearinghouse, the database of rights holder information that will support the mandatory rights holder protection mechanisms for new gTLDs.

After an unsatisfactory session during the Prague ICANN meeting, this summit was organised at short notice to address what was perceived as serious shortcomings in the proposed implementing model published by ICANN. Registries and registrars need a system that they can implement and deploy without disrupting their businesses, and without imposing costs that would discourage rights holder participation in the Clearinghouse. ICANN wants to develop a model that keeps the TMCH out of the “critical path” of domain registration, and Deloitte and IBM, who are developing the system, need a specification to build on, and a model that can become a self-sustaining entity once built.

The discussions focused on a fundamental contradiction in the TMCH specifications that are mandated by the Applicant Guidebook. During the development of the AGB, rights holders made it clear that a major concern for them was the risk of data mining if the database was released into the public domain: being able to identify new products being developed by global corporations (by tracking their registered trade marks) could threaten multi-billion dollar product development plans. This means restrictions on the distribution of the database, to minimise the chances of unauthorised access.

At the same time, in order to facilitate the Trademark Claims system, that same database has to be freely queryable by anyone wishing to register a domain name. Clearly, something’s gotta give!

A number of solutions have been proposed to try to square this circle, but ultimately there is a choice: secure, reliable, usable—pick any two. The database can be secure and reliable (held by the TMCH who can control access to it) but not usable (meeting the service levels required by registries and registrars might make it prohibitively expensive). It can be secure and usable, but not reliable, or it can be reliable and usable, but not secure (as the database is widely distributed). This is essentially a restatement of Zooko’s Triangle.

As the new gTLD program develops we are finding more issues that fell through the cracks during the development process (another example being the need to amend applications after submission). But the fundamental problems with the TMCH are much more serious: rights holder protection was probably the most controversial issue surrounding the programme, and if the protection mechanisms devised should fail, then it would be a significant blow to the credibility of ICANN as a corporation and as a community (as Kurt Pritz made clear in Brussels).

Some very smart minds are now paying close attention to this issue, trying to resolve the problems. Much of the progress made in Brussels was down to the sterling efforts of Chris Wright whose no-nonsense antipodean approach to whiteboarding the problems resolved a lot of confusion about the issue. But more feedback - especially from rights holders - is needed if we are going to build a successful platform and ensure the success of the new gTLD programme.