On Jun 2, 2010, at 11:19 PM, Shane Kerr wrote:
This protocol already exists.
Actually, multiple whois replacement protocols exist and more are coming, albeit outside of the IETF context (e.g., multiple incompatible (as I understand it) RESTful Whois implementations).
The best place to look is probably the IETF CRISP working group RFCs:
Or Whois++ (RFC 1835) or Rwhois (RFC 2167). And we shouldn't forget (like some people can't forget the taste of Tequila after overindulging) X.500 and LDAP.
The effort to implement IRIS was largely unsuccessful,
"largely"?
but not because of missing technology IMHO. It was largely because nobody cares. The problems it solves are real, but they are not any organization's primary concerns.
Right. No one knows how to make money by standardizing on how to query and display registration data and the amount of money saved by a standardized system has (to date) been insufficient to drive enough interest to fix the myriad issues with the existing whois system. As a result, folks implement whatever they feel will solve their specific problem rather than looking to solve the more generic problem, with the end result being the need for clients to know who they are querying in order to send the appropriate query and parse the resulting output. Sub-optimal, but not surprising. Regards, -drc [Speaking personally and representing no one but myself. Really.]