Summary

  • APNIC’s regional service was assembled through work performed at national network centres, with address assignment, information service and DNS handled through different institutional nodes.
  • The division evolved between 1993 proposals and the 1994 operating account; shared capacity reduced the need to place every function in one facility, while making coordination and interfaces central to the service.

The first APNIC org chart was not a building. It was a set of jobs moving between institutions. In November 1994, JPNIC described a working arrangement in which its APNIC working group played a major part in assigning addresses, Korea’s KRNIC ran the Information Service, and Australia’s AUNIC maintained APNIC DNS. The same newsletter placed those national centres under a regional APNIC layer. That is a view from one participant institution, but it reveals what “regional” meant in daily work: a request, a reference service and a name-server function did not have to sit in the same office. JPNIC’s November 1994 account

The plan changed as people tried to define it. Notes from an August 1993 IEPG meeting described an APNIC experiment with JPNIC doing network-number registration and KRNIC handling a routing registry. By January 1994, material prepared for JPNIC’s operations committee reproduced a broader proposal from APCCIRN: separate registration, directory and database services from information services. The stated reasons were to ease funding pressure and encourage cooperation among regional NICs. A criteria document was expected to explain how tasks and their interfaces could be divided. These are successive snapshots, not one frozen chart. IEPG, August 1993 · JPNIC committee material, January 1994

The timeline matters because a registry is often remembered as an authority that allocates number blocks. APNIC’s early service also needed people to answer applications, organize network information, publish records and keep DNS referrals working. RFC 1366 had described the case for regional registries, and the APNIC pilot’s charter drew on regional NICs to coordinate those practical tasks. The allocation of responsibility and the place where a request was processed were related, but they were not identical. RFC 1466 supplied the allocation guidelines the pilot aimed to apply. RFC 1466 RFC 1366 · APNIC’s history

Nor were three national desks the whole machine. Gerard Ross’s 2013 APNIC history, written by a former staff member, describes mailing lists hosted at the University of Tokyo and a staff list with 25 participants from several economies. It says participants used the list to discuss information and address requests, generally reaching rough agreement and replying within one or two working days. The retrospective also describes a workstation at the university supporting mailing lists, file archives, a network database and WHOIS; an experimental website at KRNIC; and reverse-DNS work at AUNIC.

Those details broaden the picture from a country-by-country chart to a service stitched together by people, machines and remote communication. They are retrospective descriptions, not independently audited performance figures.

The design drew on capacity that already existed in national NICs. Ross says this reduced the need to concentrate expertise at a central facility and was intended to avoid an unsustainable funding model. The contemporary January proposal likewise connected distributed operations to funding and collaboration. That rationale is credible evidence of intent; it does not quantify savings or prove the arrangement made APNIC resilient to outages. When work is shared, the interfaces become operational dependencies: a request can cross one institutional boundary for assignment, another for information and another for DNS.

The sources do not provide a complete failure or continuity record.

APNIC’s pilot began in September 1993 and formally ended in June 1994, according to APNIC’s later chronology. The work did not reduce to the pilot’s legal endpoint. A November 1994 JPNIC report still describes cross-country functions and shows APNIC address-allocation activity from April to October. The more revealing history is therefore not simply that a pilot became a regional registry. It is how regional service became executable before every task, staff member and tool occupied one institutional centre. The evidence supports a distributed operating model, while leaving its full staffing, cost and resilience unresolved. APNIC, The First 20 Years