Summary

  • Anne Lord and Randy Bush's June 2003 APNIC proposal required written policy proposals four weeks before a meeting, consensus at the Policy SIG and Members Meeting, a further public comment period, and Executive Council endorsement. Lord presented it at APNIC 16 in Seoul.
  • The Seoul room chose an eight-week post-meeting comment period, extendable when needed. Participants discussed language differences, sparse mailing-list activity, urgent decisions and whether extra time could repair weak consensus. The Members Meeting accepted the recommendation; APNIC's record dates final EC endorsement and implementation to 24 December 2003.
  • A 2014 change later let the SIG chair set the period between four and eight weeks. The story of the clock is also a story about who can keep a decision open, who judges silence, and who makes the final institutional call.

The meeting that needed a meeting rule

The item in the APNIC 16 meeting record sounded like administrative housekeeping: policy process modification. Anne Lord had brought a question with a rather larger footprint. Address policy determines how an Internet registry distributes and records number resources. If the rules governing that policy can be settled by the people who happen to be in the room, a meeting ticket starts to resemble a constitutional credential.

Lord's presentation to the Policy SIG in Seoul on 20–21 August 2003 described an existing process whose shortest path to a decision was approximately three weeks. She identified a practical defect: relevant groups could miss the chance to review a proposal before the decision was taken. The cure she presented had a calendar, a sequence of public forums and a clear final stop. This was a proposal about the path from an idea to an APNIC rule, presented through the very process it sought to change.

The revised text dated 10 June 2003 names Lord and Randy Bush as proposers. It came after discussion at APNIC 15 and on the Policy SIG mailing list. The proposal required a written paper four weeks before an Open Policy Meeting. Authors who missed that deadline could still bring an idea to the meeting for discussion; they would have to return at a later meeting for a decision. That distinction gave a late idea a microphone while keeping a late paper from becoming a surprise rule.

The draft also asked authors to state the problem, the precise change, its advantages and disadvantages, and the effect on APNIC members. These are ordinary looking headings. Together they make the cost of a proposal visible before the room expresses approval. A member who runs a network in another economy needs the change in writing, with time to ask what happens to their own operation. The proposal made that opportunity a scheduled part of the policy path.

Eight weeks, or a half-year in the waiting room?

The long argument concerned the interval after the room had spoken. The draft offered two options: eight weeks of mailing-list comment, or a period lasting until four weeks before the next Open Policy Meeting, roughly 26 weeks. Its own estimated minimum path from initial proposal to implementation was 26 weeks under the first option and 43 under the second. Those figures were the proposers' planned timetable. The meeting had to decide how much of that time served scrutiny and how much would simply make a decision age on the shelf.

The minutes of the Seoul discussion capture several pressures in one room. Some participants thought the mailing lists were quiet enough already; adding months to a silent list seemed a curious way to manufacture input. Others pointed to decisions that could affect a broad set of stakeholders across a region divided by languages. Someone suggested a four-week period. The chair proposed eight weeks with room to extend it. Representatives from other RIRs described their own process reviews. Nobody could make the calendar disappear by calling it procedure.

The vote by show of hands strongly favored the eight-week option with possible extension. The chair summarized the outcome as written proposals a month before the policy meeting, a Policy SIG finding followed by Members Meeting confirmation, eight weeks of post-meeting comments, and Executive Council confirmation of the final consensus. A participant raised a sharper question: if consensus was weak at the meeting, would more weeks of comments really mend it? The minutes say most attendees did not share that concern. It remains the most interesting objection in the record.

More time can reveal a problem; the chair still has to decide what the problem does to the proposal.

The Members Meeting then accepted the Policy SIG recommendation without objection. That act mattered, and the process continued. APNIC's prop-001 register records a comment period announced on 25 September and closed on 20 November 2003. The Executive Council endorsed the change on 24 December; APNIC marked the policy active that day and published official documentation the following February. The show of hands in August started a chain of decisions. It did not finish it.

A pause with a gatekeeper at each end

The architecture had two ambitions. A proposal should be visible before a meeting, and a room's consensus should remain open to challenge afterward. Lord and Bush's draft gave the Policy SIG chair the task of observing whether general agreement existed. After the comment period, the chair and co-chairs would decide whether substantial objections had unsettled that agreement. The Executive Council would then endorse the policy for implementation or send it back for further discussion. APNIC members could also face a formal vote if the EC chose that route under the by-laws.

This sequence widened the window for evidence and objection. It also located judgment in named roles. A member's objection posted during the eight weeks could reach the chair without requiring a flight to Seoul. A chair's assessment of whether that objection was substantial remained decisive. The EC retained a final institutional lever. The procedure gave more people a chance to be heard while keeping the question of what their words changed inside a chain of human decisions.

The regional language point deserves more than a polite nod. The minutes explicitly cite language differences as a reason to allow time for discussion. A policy paper delivered in English at a meeting can be immediately legible to some operators and only later examinable by others, after colleagues compare wording, seek advice and calculate operational effects. An eight-week period cannot guarantee anyone's participation. It gives a distant operator an actual interval in which to discover that a rule may reach their network and to say so in a public record.

APNIC's later account of the process quotes Lord recalling that older members knew informal customs while newer members could be excluded by their invisibility. That is her retrospective explanation, given years after the meeting. The 2003 paper and minutes supply the contemporaneous mechanics: a deadline, two meeting stages, a public interval, a chair's judgment and an EC endorsement. A practice that lives only in the memories of veterans favors veterans. Writing down its steps made the entrance easier to find.

The clock moved again

The eight-week period was eventually shortened. APNIC's prop-108 record shows a later proposal by Dean Pemberton and Izumi Okutani that let the Policy SIG chair choose a post-meeting comment period of four to eight weeks. APNIC announced implementation on 19 February 2014. The original clock was durable enough to become part of the institution and adjustable enough to be revisited by that institution.

The 2014 change sharpens the 2003 story. A fixed eight weeks had given the second audience a predictable minimum span; a four-to-eight-week range gave the chair more control over the schedule. Either design still puts weight on the quality of the text submitted before the meeting, the visibility of the list, and the willingness of participants to explain concrete effects. A clock can hold a door open. Someone has to walk through it with an argument.

Lord's contribution in Seoul was to put that door into the written route from proposal to policy. Bush shared the draft; the room chose the interval; the Members Meeting and EC gave it institutional force. The most memorable part of the debate may be its self-reference: the people deciding how long people should get to debate policy had to debate how long that debate should last. They chose eight weeks. Then they waited through them.