Summary
- In 1992, could Internet registration move closer to regional networks without splitting the common address system? RFC 1366 answered with eight proposed Class C blocks and one retained root.
- It kept the Internet Registry as the root and default, retained all Class A and B space centrally and proposed eight distributable blocks from the then Class C range.
- The proposal describes coordination criteria and a staged plan; its limited group support is not evidence of a global user mandate or regional ownership of addresses.
A registry closer to the network
A new network could join the Internet only after it had been assigned numbers. In 1992, as networks crossed borders, that practical step raised a larger question: who would register those numbers, and where? RFC 1366 proposed regional registries to bring the work closer to operators and subscribers, while keeping one central root and a shared allocation framework. Its plan set out eight Class C blocks; the Internet Registry would remain the default and retain every Class A and B network. The RFC records a proposed route into a larger Internet, not proof that later operations followed every boundary it drew.
The document’s stated case for regional registration was practical: a registry located nearer to its users might better serve local languages and customs. It named Europe, Central and South America, and the Pacific Rim as places where the Internet had become mature enough to consider a regional organization. The design did not replace the central Internet Registry. The central registry would remain the root, retain unallocated responsibility for other regions and be ready to serve a network directly if referral to a regional registry failed.
The boundary was written into the address plan
RFC 1366’s June 1992 table counted 44,014 of 2,097,151 Class C networks as allocated, or 2%. It reported 38% of Class A and 45% of Class B networks allocated. These A/B/C labels belong to the classful network-number plan of that period, not to today’s CIDR prefixes. Because A and B numbers were considered limited, the proposal kept all of them in the Internet Registry. Only Class C space would be divided for regional delegation.
Eight equal-sized blocks were proposed, each containing 131,072 addresses, approximately 6% of the Class C space. The proposed map paired some ranges with Europe, North America, the Pacific Rim and Central/South America; it also left ranges labelled “Others” and one multi-regional block. That geography organized service and aggregation. The RFC did not say that addresses belonged to the region whose name appeared beside a range.
A regional registry was expected to be unbiased and recognized by providers and subscribers in its area. Selection criteria included local networking authorities’ legitimacy, an established organization, a commitment to stable service, adherence to IANA and Internet Registry guidelines, and coordination on sub-allocation. RFC 1366 also proposed one registry per geographic area at that level. These rules made delegation conditional and bounded: a regional body handled registration within a plan set by the central registry.
A plan, not a plebiscite
The RFC was informational, not an Internet standard. Its abstract records “general consensus” among the Federal Engineering Task Force, Federal Networking Council, IEPG co-chairs and RIPE to support its recommendations. That is evidence of support by named technical and administrative groups, not a vote by Internet users or proof that every network accepted the scheme.
A companion RFC 1367 put dates against the proposal: candidate registries could be considered from 31 October 1992; the Internet Registry was expected to begin Class C allocations under the plan on 15 April 1993; an aggregation mechanism was anticipated by 6 June. The memo called this a draft schedule and invited open discussion. Those targets describe what planners expected, not a verified record that each milestone occurred as written.
The European case had already moved ahead institutionally. RIPE’s 1990 proposal called for a network coordination centre with a regional registry role, and its first activity plan followed in May 1991. The RIPE NCC began operating in April 1992, months before RFC 1366 appeared. The RFC therefore did not create the first regional registry; it formalized a broader allocation design around a service model already being built.
The Pacific Rim block has one documented sequel. APNIC’s own history says a regional NIC plan was accepted at the 1993 APCCIRN meeting; its pilot was set for September 1993 through June 1994. In April 1994, IANA delegated 202/8 and 203/8 to APNIC—the same pair the RFC table marked for the Pacific Rim. This shows one later allocation path, not that every planned block or date was realized.
The important historical distinction is between a ledger’s reach and its authority. In 1992, distribution could make registration more responsive while preserving one root record, a central fallback and a common allocation scheme. The proposal’s own limits—especially the retained Class A and B space—show that “regional registry” did not yet mean a region controlled all numbers used within it.
Sources
- RFC 1174 — IAB Recommended Policy on Distributing Internet Identifier Assignment (1990)
- RFC 1366 — Guidelines for Management of IP Address Space (1992)
- RFC 1367 — Schedule for IP Address Space Management Guidelines (1992)
- RIPE NCC — The RIPE Registry
- RIPE-039 — Procedures To Set Up The RIPE NCC (1991)
- APNIC — The First 20 Years
Member Briefing
Deeper Profile Context
Sign in with the right membership level to unlock the full briefing and source notes.
Only for Strategic Circle
Strategic Circle
Open to all readers. Unlock profile briefings after joining and signing in.
Join Strategic CircleOnly for Leadership Alliance
Leadership Alliance
For qualified IP-asset owners and management; sign in to unlock alliance briefings.
Join Leadership Alliance
