Summary

  • The APNIC 62 NIR SIG deck proposes expanding a long-standing information-sharing charter into best-practice support, policy implementation, view consolidation and possible referral to the APNIC Executive Council.
  • Its survey slide lists four materially different questions and says feedback came from seven NIRs, but shows only one pie chart: 86% support, 14% neutral and 0% opposition.
  • The chart is not attributed to any question and no aggregation method is given; it therefore cannot establish which part of the proposed mandate received the reported support.
  • A question-level consultation receipt would preserve respondent privacy and the later consensus process while making each support claim auditable.

Four propositions enter; one number leaves

The charter review scheduled for APNIC 62 begins from an unusually legible comparison. APNIC’s current NIR SIG charter is one sentence. The forum shares information about the operations, policies and procedures of National Internet Registries and promotes cooperation among those registries and the APNIC Secretariat.

The proposed text adds several jobs. It would make the SIG a place for best current practices and effective policy implementation. It would strengthen networking among NIRs. Most consequentially, it would identify common issues, consolidate views and recommendations from NIRs and NIR members, and, where appropriate, communicate them to the APNIC Executive Council for consideration.

That is not merely a stylistic refresh. Information can be shared without deciding what the group thinks. Consolidating views requires a rule for moving from several positions to a collective output. Sending that output to a governing council adds an institutional recipient. The proposed wording does not bind the EC, but it creates a channel through which a summary may acquire authority from the name of the forum.

The survey slide is therefore more than conference decoration. It asks four questions: whether the additions on best practices and policy implementation are welcome; whether collaboration language should be added; whether the SIG should consolidate views and send them to the EC; and whether respondents support the revision overall.

To the right is one chart. It reports 86% support, 14% neutral and 0% “do not support”. The slide does not label the distribution as the answer to Question 1, Question 2, Question 3 or Question 4. It does not show four distributions or describe a method for combining them.

The arithmetic tempts the reader. APNIC says there are seven NIRs in its region, and the slide says feedback came from seven NIRs. If each supplied one equally weighted answer, 86% and 14% would be consistent, after rounding, with six supportive responses and one neutral response. But the deck does not publish those counts, confirm equal weighting or identify the question. That calculation is an inference, not a result the source reports.

The missing label matters because the questions are not substitutes

A registry could support practical cooperation while resisting a collective advocacy role. Another could favour referral to the EC but think the proposed charter too broad overall. A third could accept the whole revision yet want safeguards around minority views. Those are not exotic distinctions. They follow from the four questions APNIC’s own deck chose to ask.

One distribution cannot be carried across all four propositions without a stated rule. Nor can overall support safely be treated as support for the most consequential clause. A package can win approval despite reservations about one component; a component can command support while the full package does not.

This is an attribution problem, not a complaint that seven respondents are statistically too few. APNIC’s current public page identifies exactly seven NIRs: APJII, CNNIC, IRINN, JPNIC, KISA, TWNIC and VNNIC. If the survey reached all seven, it may have covered the complete institutional population it intended to ask. A census can still produce an ambiguous publication if the answer is not joined to its question.

It is also not evidence that four result sets do not exist. A slide is a presentation surface, not a database audit. Detailed counts may be available to the chairs, may be explained during the session, or may appear in minutes later. The defensible statement is narrower: the deck published before the scheduled session does not let an outside reader make the attribution.

The strongest defence is process after the slide

The chart need not decide anything. APNIC’s SIG Guidelines define SIGs as open public forums, not committees limited to incumbent registries. The deck ends with an invitation for views. More importantly, the active Guidelines say an altered SIG charter requires consensus at the AGM or APNIC Member Meeting. Their definition of consensus is not 86%. It is a condition in which everyone consents, even when the decision is not everyone’s first preference.

That sequence can correct a thin survey display. A compact chart may help a live audience grasp a broad direction. Naming individual NIR answers in a group of seven could reduce candour. Chairs may reasonably protect respondent identity, and the proposed words “where appropriate” and “for consideration” preserve discretion at both ends of the referral.

These are serious protections. They are also reasons to ask for better aggregate reporting rather than names. Privacy does not require mixing four questions into one result. A later consensus call is easier to understand when participants know which proposition drew neutrality or reservation.

A question-level receipt is enough

The remedy does not require a public ballot history. For each question, publish the exact wording and charter version, the number invited and responding, counts for support, neutral, oppose, abstain or missing, and the weighting and rounding rule. A short privacy-safe note can summarize material reservations without identifying the NIR that raised them.

The receipt should then record what happened at the SIG and at the AGM or AMM: retained wording, amendments, consensus finding, date and responsible chair. If the revised charter is adopted and the SIG later sends a consolidated view to the EC, a second receipt should identify the issue, input window, consolidation method, minority treatment, sender, date, EC acknowledgement and eventual disposition.

None of this turns the survey into a referendum. It does not give NIRs power to bind the wider APNIC community, or the SIG power to bind the EC. It makes the weaker and more useful claim: this reported answer belongs to this question, under this version, with this denominator.

The APNIC 62 schedule placed the NIR SIG session at 16:30 IST on 8 September. The source package for this article was frozen before that session began. No conclusion here assumes that the presentation occurred, that participants endorsed the revision or that the charter was adopted. The timely fact is the publication of a proposal and its evidence surface.

The NIR SIG’s proposed new role is to turn dispersed views into an intelligible institutional message. Its charter review should demonstrate that craft. Four questions do not need one bigger number. They need four attributable answers.

Sources