Summary

  • RFC 1591 said IANA would establish the Internet DNS Names Review Board (IDNB) to decide unresolved disputes over who should manage a top-level domain, and that its decisions would be binding.
  • In 2017 and 2024, ccNSO records said the IDNB had never been established. A different, nonbinding ccTLD review policy was adopted in 2025; as of the ICANN86 report in May 2026, its status was implementation planning.

A paragraph in a 1994 memo reads like the end of an argument. If parties competing to manage a top-level domain could not agree, the Internet DNS Names Review Board would act as a review panel. Its decisions would be binding. The sentence sat inside RFC 1591, Jon Postel’s account of DNS structure and delegation, between a preference for agreement and a warning that a prospective manager might not represent a whole country.

The board is the surprise. RFC 1591 was not simply a checklist for choosing a technically capable registry. It sketched a way to manage disagreement over who would run a domain, including a body that could issue a final decision. But later ccNSO records say that the IDNB was never established by IANA or any other entity. The text described an institution that, according to the retrospective record, did not come into being.

The surrounding rules matter. RFC 1591 marked itself “Informational” and expressly said it did not specify an Internet standard. It described IANA as responsible for overall DNS coordination, while the central Internet Registry—then the InterNIC—handled much of the day-to-day administration. For each domain, it wanted a designated manager with working name servers, local administrative contact for a country-code domain, and the ability to operate competently.

The memo framed that role as trusteeship. A manager served both the national community and the global Internet community; arguments over ownership or rights to a domain were called the wrong concern. The proper questions, it said, were responsibility and service. It required nondiscriminatory treatment of applicants, including no preference for customers of a network provider related to the manager. A subdomain could operate its own name servers, provided the information was accurate. These were concrete constraints on how delegated authority should be exercised.

Then came the consent language. “Significantly interested parties” should agree that the designated manager was appropriate. IANA should try to get contenders to agreement and generally avoid changing an established delegation unless all contending parties agreed, except where the manager had substantially misbehaved. Interested parties should have “some voice” in choosing a manager. Yet the memo also allowed IANA and the central registry to delegate only part of a new TLD when contenders could not agree or when an applicant could not represent the whole country.

It ended that branch of the procedure by naming the IDNB as the review panel, with binding decisions.

That sequence joined participation to a decision mechanism without making them the same thing. Interested parties were to be heard; the memo did not say their attendance itself settled a dispute. A designated manager carried operational duties, while IANA retained the coordinating role described in the memo. The IDNB was supposed to provide the adjudicative step when agreement failed. Whether that step existed in practice is therefore not a footnote. It determines whether the written route had a working destination.

A June 1996 Internet-Draft by Postel carried the IDNB into a different debate. It proposed a policy and control structure for international top-level domains such as COM, ORG and NET, and expanded the board’s role to include technical appeals about that proposal’s process. The draft also divided responsibilities among the IANA, IETF, ISOC and an ad hoc group. But it remained an individual Internet-Draft: the Datatracker says it was not endorsed by the IETF and had no formal standing in the standards process. It expired in December 1996.

It is evidence that the board concept was reused in a proposal, not evidence that the proposed body was constituted or that the design became policy.

The absence became explicit in later institutional records. A ccNSO charter working group adopted in March 2017 said the IDNB had never been established and observed that decisions involving ccTLD delegation, transfer and revocation lacked a review or appeal mechanism. The September 2024 ccNSO final report repeated the point and said such decisions were not then subject to independent review. These are later records describing the gap; they do not prove every dispute went unheard or establish what happened in any individual delegation case.

There is a second ending, and it is not the IDNB arriving late. On 8 June 2025, the ICANN Board adopted a separate ccTLD Review Mechanism Policy developed through the ccNSO. The public announcement described an independent review path for specific IANA Function Operator decisions, but said the outcome would not be binding. The ICANN86 Policy Outlook Report, dated 14 May 2026, listed the policy’s status as “Implementation planning.” The scope, institutional setting and force of this mechanism differ from the 1994 board envisaged in RFC 1591. It should not be treated as that board’s belated establishment.

The historical distinction is modest but consequential. RFC 1591’s text recognized that affected parties had a stake in choosing a manager, placed service and fair treatment above ownership claims, and contemplated a binding review body when agreement failed. Yet a named panel was not an operating institution. Later efforts to design a review mechanism had to start from that absence and work through a separate policy process. Participation can provide knowledge, objections and evidence. A review route requires something more concrete: a body, a scope, a procedure and a defined consequence.

In the DNS record, those pieces did not arrive together.

Sources