Summary

  • AFPUB-2026-GEN-001-DRAFT01 would formalise roles and procedures for AFRINIC's Policy Development Working Group, including participation, co-chair administration, consensus and institutional fallback.
  • At AFRINIC 37 on 24 June 2026, the authors said the proposal was not ready to seek consensus; the meeting record says it returned to the mailing list and AFRINIC still lists it as under discussion.
  • The durable design test is whether every legitimacy-sensitive handoff leaves an auditable record of the rule, actor, input, reasons, dissent, deadline and review path.

A pause can be evidence of process working

The most revealing fact about AFRINIC's proposed working-group rules is not a clause. It is what did not happen on 24 June. The authors used the AFRINIC 37 meeting to explain the proposal and gather feedback, while telling the meeting that the text was not ready for a consensus decision.

That restraint matters. A public-policy meeting can create pressure to turn presentation into decision, especially when participants have travelled, agendas are fixed and institutional uncertainty makes closure feel valuable. Declining to seek consensus preserves the distinction between explaining a design and asking the working group to authorize it.

AFRINIC's current proposal index records draft 1, submitted on 20 May 2026, as under discussion. The meeting minutes say it was sent back to the mailing list. Those states are consistent: the proposal remains alive, but it is not adopted, ratified or implemented.

This chronology—submission on 20 May, circulation to the RPD list on 25 May, presentation without a consensus request on 24 June, and continued under-discussion status at the reporting freeze—should remain visible beside any future version. A single status badge cannot carry that history.

Written procedure is only the first layer

The proposal responds to a real governance problem. The authors argue that the PDWG has relied on a thin baseline and accumulated practices that different participants interpret differently. When unwritten practice is invoked, one side can call it established custom while another calls it an invention. Writing down roles and procedures can reduce that ambiguity.

But codification does not automatically make a decision legitimate. A written rule can still leave unclear who supplied the decisive input, which version governed, whether the deadline was met, how dissent was treated or which institution took responsibility for the next step.

The draft reaches across a wide control surface: who may participate, how co-chairs are selected or replaced, how the working group organises itself, how consensus is assessed, what happens when consensus cannot be reached, how institutional actors respond and how decisions can be challenged. These are not clerical details. They determine who can move a proposal from discussion toward rules that affect number-resource administration.

That is why the implementation object should be more than a revised policy manual. AFRINIC needs a decision ledger for the process itself.

Record every transfer of authority

A useful ledger would treat each transition as a separate event. Submission should identify the received text, author, timestamp and assigned proposal ID. Publication should bind the public page to an exact version. Meeting eligibility should record the applicable deadline and decision. Consensus assessment should identify the discussion window, evidence considered, objections, chair conclusion and reasons. Where the draft assigns a fallback to the Number Council—such as appointing an interim co-chair after the working group fails to reach consensus—the ledger should identify the trigger, deadline, appointee and reasons.

Ratification should record the Board's action and legal or procedural basis. Implementation should identify the effective text and deployment date.

The links between those events matter as much as the entries. A later draft should not silently inherit support expressed for an earlier version. A meeting presentation should not be mistaken for a consensus call. A chair conclusion should not be confused with Board ratification. Ratification should not be reported as implementation before the operative systems and manual have changed.

The ledger need not expose private information or turn every mailing-list message into a vote. It should expose enough structured evidence for a participant to reconstruct why the process changed state.

Participation rules need a purpose test

The RPD discussion illustrates why eligibility language requires more than a neat threshold. One participant objected that requiring recent in-person meeting participation would be unrealistic in a region where travel, cost and connectivity are uneven. The same comment opposed consensus appointment of co-chairs as too subjective and argued that the Board should retain a reasoned ability not to ratify.

Those are participant views, not findings by AFRINIC. They nevertheless identify three distinct design questions. First, does an eligibility criterion measure relevant contribution or merely physical presence? Second, can a selection rule be applied consistently when the same people also interpret consensus? Third, when an institutional actor declines a community outcome, must it publish reasons and trigger a defined review path?

Each rule should be tested against its stated purpose. If the purpose is demonstrated engagement, the evidence could include sustained mailing-list work, remote participation, drafting, issue analysis and meeting contributions—not only travel. If the purpose is trusted administration, the selection record should show candidates, eligibility, conflicts, decision method and any unresolved objection. If the purpose is a bounded Board safeguard, the rule should require reasons, a deadline and a return path rather than indefinite silence.

Separate consensus from institutional completion

Rough consensus is not a count of supportive messages, but neither can it be a conclusion without an evidentiary trail. A chair should be able to explain the substantial objections, how the current text addresses them and why remaining disagreement does or does not prevent progress. Participants should be able to see which version received that assessment.

The next handoff is different. Board ratification and staff implementation involve institutional duties beyond the working group's discussion. Treating those steps as automatic can obscure legal, operational or resource constraints. Treating them as open-ended discretion can nullify the policy process.

The remedy is not to collapse the institutions into one decision. It is to make the boundaries explicit. The working group owns its consensus record. The Board owns its ratification act and reasons. Staff owns an implementation plan, dependency record and effective-date evidence. An appeal body owns the review of the decision actually challenged, not a generalized re-run of the entire debate.

The next draft should make state reconstructable

AFRINIC's 24 June pause gives the authors room to improve the text before anyone treats it as authority. The most important improvement would be a common evidence model across its procedures.

For every transition, the next draft should answer seven questions: What exact text was before the decision-maker? Who had authority to act? What inputs and objections were considered? Which rule and deadline applied? What decision was made and why? What state followed? Who can review or reverse it, and on what record?

These questions do not predetermine the correct balance among participants, co-chairs, the Number Council, the Board and staff. They make the balance inspectable. That is the difference between written procedure and accountable procedure.

The proposal remains under discussion. Its current value is therefore not that it has settled AFRINIC's governance, but that it exposes the handoffs that must be settled before the working group can rely on them.

Sources