Thick Whois
Introduction: what are the implications of a 'thin' Registry possibly becoming authoritative for registrant Whois data following the transition from a thin-registry model to a thick-registry model. The Working Group should consider the term "authoritative" in both the technical (the repository of the authoritative data) and policy (who has authority over the data) meanings of the word when considering this issue.
Volker: Registrar -- dispute resolution providers come to the Registrar to confirm what they learn from Registry
The registrar is the responsible party under the UDRP
Different people may be in charge of the domain -- the manager of the account and the person who is listed as the Registrant of Record
+ - RAA -- Registrar needs to be able to tell who the reseller is -- entirely voluntary on the part of the registrar
Volker/Frederic -- yes
May serve as a backup for a registry crash
Frederic -- we have to manage the system -- easier with fewer TLDs
+ - So this may be a non- issue from our standpoint -- since Registrars already escrow thin TLDs and presumably would continue to do so
Introduction: in the event of a Registrar business or technical failure, it could be beneficial to ICANN and registrants to have the full set of domain registration contact data stored by four organizations (the Registry, the Registry's escrow agent, the Registrar, and the Registrar's escrow agent), which would be the case in a 'thick' registry.
• Current Registry and Registrar Escrow provider Iron Mountain (suggested by Volker Greimann, RrSG)
Introduction: 'thick' Whois might obviate the need for the registrar escrow program and attendant expenses to ICANN and registrars.
Introduction: what would be the impact on the registry and registrar WHOIS and EPP systems for those Registries currently operating a thin registry, both in the migration phase to 'thick' WHOIS as well as ongoing operations?
Introduction: a 'thick' Registry can dictate the labeling and display of Whois information to be sure the information is easy to parse, and all Registrars/clients would have to display it accordingly. This could be considered a benefit but also a potential cost. This might also be a benefit in the context of internationalized registration data as even with the use of different scripts, uniform data collection and display standards could be applied.
Introduction: is the provision of Whois information at the registry level under the 'thick' Whois model more effective and cost-effective than a 'thin' model in protecting consumers and users of Whois data and intellectual property owners?
Introduction: how would 'thick' Whois affect privacy and data protection, also taking into account the involvement of different jurisdictions with different laws and legislation with regard to data privacy as well as possible cross border transfers of registrant data?
• Dr. Joanna Kulesza, Faculty of Law and Administration, University of Lodz (Suggested by Roy Balleste, NCUC)
Introduction: what would be the impact on competition in registry services should all Registries be required to provide Whois service using the 'thick' Whois model – would there be more, less or no difference with regard to competition in registry services?
• Need to look at survey and sales data for both kinds of registries (suggested by Jonathan Zuck)
Introduction: What, if anything, are the potential impacts on the providers of third-party WHOIS-related applications if 'thick' WHOIS is required for all gtLDs?
Introduction: 'thick' Whois could make the requirement for Registrars to maintain Port 43 Whois access redundant.
Introduction: what are the cost implications of a transition to 'thick' Whois for Registries, Registrars, registrants and other parties for all gTLDs? Conversely, what are the cost implications to Registries, Registrars, registrants and other parties if no transition is mandated?
A benefit would be that in a thick registry one could develop a secure method for a gaining registrar to gain access to the registrant contact information. Currently there is no standard means for the secure exchange of registrant details in a thin registry. In this scenario, disputes between the registrant and admin contact could be reduced, as the registrant would become the ultimate approver of a transfer.
Centralized databases operated under a single administrator that sets conventions and standards for submission and display, archival/restoration and security have proven easier to manage. By contrast, a thin Whois model is a decentralized repository6. Registrars set their own conventions and standards for submission and display, archival/restoration and security registrant information. Today, for example, Whois data submission and display conventions vary among registrars. The thin model is thus criticized for introducing variability among Whois services, which can be problematic for legitimate forms of automation.
A thick Whois model offers attractive archival and restoration properties. If a registrar were to go out of business or experience long-term technical failures rendering them unable to provide service, registries maintaining thick Whois have all the registrant information at hand and could transfer the registrations to a different (or temporary) registrar so that registrants could continue to manage their domain names.
A thick Whois model also reduces the degree of variability in display formats.
a thick registry is better positioned to take measures to analyze and improve data quality since it has all the data at hand.
the extensible provisioning protocol (EPP) was not designed to handle the extensive updates every time a registrar makes changes to the Whois record.
Proponents of requiring thick Whois argue that being able to access the thick data at both the registry and the registrar level will ensure greater accessibility of the data. The draft report of the Implementation Recommendations Team put together by ICANN's Intellectual Property Constituency stated "the IRT believes that the provision of WHOIS information at the registry level under the Thick WHOIS model is essential to the cost-effective protection of consumers and intellectual property owners."
+ - There are at least two scenarios in which the additional option of retrieving the data at the registry would be valuable:
1. Where the registrar Whois service might be experiencing a short- or long-term outage (in violation of the registrar's accreditation agreement), and
2. Where the registrar has implemented strong (or sometimes overly-defensive) measures to prevent large-scale automated harvesting of registrar data.
Also, in the event of a registrar business or technical failure, it could be beneficial to ICANN and registrants to have the full set of domain registration contact data stored by four organizations (the registry, the registry's escrow agent, the registrar, and the registrar's escrow agent) instead of just two organizations (the registrar and the registrar's escrow agent).
IPC notes that ‘thick’ Whois is not the only answer to enhance data accuracy, but ‘it is one step that, along with a number of other changes, could move us toward a solution’ pointing to a recent study conducted by NORC at the request of ICANN which demonstrates that a ‘thick’ Whois model provides more accessible and accurate contact data.
+ - IPC is of the view that centralized access to Whois data in a ‘thick’ model would bring a number of advantages including
In the opinion of B, ‘thin’ Whois ‘helps maintain Whois data inaccuracy’, while VS points out that ‘thick’ Whois does not necessarily improve data accuracy as ‘the registry only displays data that is provided by the registrar and the registry has no way of validating registrant data’.
+ - Pay particular attention to the degree to which we look at how all thick-Whois registries are managed, vs questions that bear on transitioning all registries to thick.