The Future of WHOIS
APNIC is working on a new standarized WHOIS protocol. APNIC Technical Director Byron Ellacott regularly attends Internet Engineering Task Force (IETF) meetings.
Byron Ellacott, APNIC Technical Director
The WHOIS protocol is used to discover registration information about network resources – it is the protocol behind the whois service, which is a more general term for such registration database queries. Originally created in the 1980s to discover network operators on ARPANET, the WHOIS protocol now has additional purposes, including checking for availability of names, law enforcement investigation, and trademark research. There are now many more stakeholders interested in WHOIS than were originally envisioned, and the protocol is groaning under the weight of that additional responsibility.
Today, WHOIS offers four main problems to publishers and consumers of registry data. The WHOIS standard is very brief, and does not specify query types, output formats, or error messages, and each whois service operator has made those decisions independently. There is also no support for internationalization, which has become a much more pressing problem with the advent of internationalized domain names (IDNs). The protocol was designed for a single server answering all queries, and has no mechanism to cope with today’s distribution of registrations and authorities. There is also no support for authentication mechanisms, again leaving the challenge of different classes of users for each operator to resolve.
During the past 30 years, several attempts have been made to address these shortcomings. In 1993, an attempt was made to use X.500 Directory Services to replace WHOIS and other information access protocols, but no end product was created as a result. In 1994, Network Solutions (now VeriSign) developed Referral Whois (RWhois), which saw some limited success, but never replaced WHOIS. In 1995, the IETF created Whois++ (RFC 1834), but this protocol was never deployed. In 2005, the IETF created the Internet Registry Information Service (IRIS, RFC 3981), which is in use among registries and registrars today, but is too complex for a general purpose information access protocol.
Recently, a successor to WHOIS has emerged. Two different web services approaches, one by ARIN and one by the RIPE NCC (the Regional Internet Registries for North America and Europe and the Middle East, respectively), have both been successfully deployed to supplement the traditional WHOIS services run by those operators. The IETF is now in the process of standardizing that success, through the Web Extensible Internet Registration Data Service, or WEIRDS. Operators and consumers of a variety of whois services are collaborating to produce a protocol that is almost as simple as the original WHOIS, while addressing the its shortcomings.
The WEIRDS protocols will describe how to ask for, and respond with, registration data. They will not refer to what that data should be, or make any recommendations on new data to be included. The process is a representational transformation, like creating an e-book from an old printed novel. The data does not change, but it becomes more easily accessible to a growing number of users. Current discussions active about privacy, fine-grained data, and accuracy in the WHOIS are separate to the WEIRDS effort. These are policy decisions, and the engineers collaborating on WEIRDS are ensuring the new protocol leaves as many options as possible.
The WEIRDS working group plans to create several standards documents, including a base protocol specification and separate query and response specifications for domain names and Internet number resources. The project works in parallel on the two forms of whois service, focusing on delivering a protocol that leaves as many options as possible to the final operator of the service. Documents are expected to enter the IETF publication process from November 2012, with the documents that specify how a WEIRDS server for an Internet numbers registry should operate targeted for March 2013.
While the IETF produces a new protocol, momentum is also gathering in the policy arena, where questions about contract terms between ICANN and Internet registries are being discussed. These discussions also include concerns about data accuracy, privacy and the use of WHOIS for law enforcement purposes. It's important to distinguish the bottom-up policy processes, which are agreed to and applied evenly to all parties, in contrast to protocol discussions that are held at a higher level.