Summary

  • AFRINIC published AFPUB-2020-GEN-004-DRAFT01 on 13 August 2020; the record then identifies AFPUB-2020-GEN-004-DRAFT02 as the next version, published in September, and records a declaration of consensus at the September process stage.
  • The available material proves a D1-to-D2 transition and a new procedural status. It does not provide a complete authenticated clause-by-clause redline, so it cannot support claims that any particular clause was added, deleted, narrowed, enlarged or left unchanged.
  • A reconstructible consensus judgement needs an objection ledger tied to the exact Draft 2 text: the proposition, objection, evidence, response, disposition, reason, resulting text and unresolved state. The record available here does not reproduce that complete chain or disclose a denominator for affected-party review.
  • Consensus, corporate competence, contract and public authority are separate tests. Neither a PDP, a consensus declaration, Board status nor an RSA can bootstrap another instrument into unlimited authority.
  • A legitimate Board risk interface is possible, but only if it is confined to named corporate-law, contractual, security, feasibility and continuity risks, supported by evidence and reasons, time-bounded, reviewable and separated from drafting and execution.
  • NRS, Lu Heng, LARUS and BTW illuminate different parts of the problem: member protection, the bookkeeper boundary, operator continuity and disciplined verification. None supplies AFRINIC with powers it did not otherwise possess.

L3 — The version the September call actually moved

The decisive word in September 2020 was not Board. It was consensus. At the point when the co-chairs used it, the word altered the procedural position of AFPUB-2020-GEN-004-DRAFT02, Board Prerogatives on the PDP, Draft 2. It recorded a judgement about that draft within AFRINIC’s Policy Development Process. The moment matters because a procedural label can look deceptively conclusive. It can compress a sequence of texts, interventions, objections and judgements into a single public status. Yet the status answers only the question allocated to that stage. It does not, by the force of its vocabulary, settle the lawful reach of the underlying proposal.

AFRINIC 32’s public-policy meeting took place on 16 September 2020. The available source record identifies the institutional act as publication of Draft 2 and a declaration of consensus during that September process. Those are the events that can safely be placed together: there was an identified second draft, and that draft acquired a stated procedural status. Nothing in that formulation requires the reader to assume what occurred beyond the boundary of the declaration. Nor does it require an inference that every person affected by the proposal agreed with its substance, or even reviewed precisely the same text.

That distinction is more than lawyerly caution. Version control determines what a community supposedly accepted. If a proposition is controversial, the difference between accepting an objective, accepting a drafting direction and accepting exact operative words can be substantial. A declaration without a securely identified text leaves readers unable to test which of those objects was actually assessed. A declaration tied to an exact text is much more useful, but even then it remains a process judgement. It says that the decision-maker charged with assessing consensus reached a stated conclusion under the applicable process.

It does not automatically say that the subject matter lay within every institution’s competence.

Draft 1 is the baseline, not the subject

The version history begins with a fact that is both useful and limited. AFPUB-2020-GEN-004-DRAFT01, Board Prerogatives on the PDP, Draft 1, was published on 13 August 2020. In the available record, its baseline architecture made several categories of possible Board intervention explicit: initiating or varying policy text, suspending a policy, declining ratification, or requiring policy action. Those categories explain why the proposal raised a structural question. They describe a contemplated interface between policy development and a corporate Board, rather than an ordinary allocation rule inside the registry ledger.

But Draft 1 does not own this story. Its categories are a starting coordinate, not a licence to reconstruct the second draft from memory, implication or later descriptions. Draft 2 is the next identified version, and the record dates its publication only to September 2020. The exact day of publication is not established here. More importantly, the available material does not supply a complete authenticated line-by-line comparison between D1 and D2. A responsible account must therefore resist verbs that pretend to know what the comparison would show.

It cannot say that D2 added, removed, preserved, tightened, relaxed, expanded or narrowed a particular clause.

This prohibition is not pedantry about documents. A clause-level claim can change the political and legal meaning of the event. If one says, without a reliable redline, that an objection caused a power to be narrowed, one has invented both a textual development and a successful institutional response. If one says that the controversial language survived unchanged, one has invented a different story: that the process heard a challenge and proceeded regardless. Both versions may sound plausible. Neither is evidence.

The only supported account is that a first draft provided the enumerated baseline, a second identified draft existed, and the September declaration attached a consensus status to the second version.

The discipline also protects critics from overclaiming. It would be easy to turn uncertainty into an accusation that the drafting process concealed changes. The absence of a complete authenticated redline in the material considered here does not establish concealment, bad faith or procedural evasion. It establishes an evidence limit. A redline might exist elsewhere; this article does not assert that it does or does not. The correct response is to name the missing comparison and explain why it would matter, not to supply an imagined substitute.

Three objects must not be collapsed

The September moment contains at least three distinct objects. The first is the proposal family represented by the D1 and D2 identifiers. The second is the exact Draft 2 text placed before the process. The third is the co-chairs’ consensus assessment about that text. Public discussion often collapses all three into a sentence such as “the community approved Board prerogatives”. That sentence is too coarse to be reliable. It substitutes a general constituency for identified participants, “approved” for a particular consensus procedure, and an abstract policy label for the version actually assessed.

The first object supplies continuity: it tells readers that D2 belonged to the same named proposal sequence as D1. The second supplies the operative language: without it, no one can locate a disputed proposition or determine the scope of the drafting. The third supplies a process decision: it explains why the draft could move within the PDP. Each object has a different evidential burden. A proposal identifier can establish custody and version lineage. A preserved text can establish words. A meeting record can establish that a declaration was made.

None, alone or together, proves that a private institution thereby obtained a new category of legal authority.

This is where the phrase “final intervention clauses and objections treated as resolved” must be handled carefully. It identifies the subject of the consensus judgement in the available record. It does not disclose, by itself, every intervention clause, every objection or the reasoning by which each objection was considered resolved. “Treated as resolved” reports a procedural outcome. It is not a substitute for the chain of evidence that would let an independent reader reconstruct the judgement.

A reliable account should therefore formulate the September finding narrowly: the co-chairs declared consensus on the identified Draft 2 in the September process. That declaration was capable of advancing the text under the private process if the process’s rules were satisfied. It did not turn the proposition’s title into a grant of power. It did not convert an assessment of objections into assent by all affected resource holders. It did not establish the content of an absent redline. And it did not resolve the separate sources of corporate, contractual or public authority.

What the record proves, and what it leaves open

The positive evidence can be stated without hesitation. Draft 1 was published on 13 August 2020. It offered the baseline categories of contemplated intervention. Draft 2 was the next identified version and was published in September 2020, although the exact publication day is not established by the material used here. AFRINIC 32’s public-policy meeting took place on 16 September. A September consensus declaration was made about Draft 2. These facts support a version transition and a new procedural status.

The corresponding unknowns are equally important. The record considered here does not contain a complete authenticated D1-to-D2 redline. It does not reproduce a complete ledger linking each material objection to its evidence, an author response, a co-chair disposition, a reason and a location in the final text. It does not disclose a denominator showing how many affected members, operators or number-resource holders reviewed the exact Draft 2 wording. It does not show that the consensus declaration itself supplied legal authority for the proposed intervention categories.

And this article draws no conclusion about any stage after that September declaration.

Unknowns should remain grammatically visible. “The record does not show” is different from “it did not happen”. The first describes the limit of the evidence available for a claim; the second asserts a fact about the world. That difference is essential in governance reporting, where incomplete minutes or summary records may preserve an institutional act without preserving every input to it. The absence of a denominator does not prove that participation was small. The absence of a complete objection ledger does not prove that no objections were heard or that all objections were ignored.

The absence of a redline does not prove that the versions were identical or radically different.

This approach also prevents uncertainty from being used in the opposite direction. An institutional summary cannot be treated as though it contains detail that it does not reproduce. A declaration of consensus cannot stand in for the objections, responses and reasons that would make the declaration independently testable. Neither scepticism nor deference is a method of filling gaps. The evidence boundary applies both ways.

The denominator cannot be inferred from the room

Policy meetings and mailing lists can produce valuable technical knowledge. Participants may have direct operational experience, and objections can expose implementation risks that formal consultation would otherwise miss. But the presence of useful participation does not reveal the size or composition of the affected population. A meeting attendance figure, were one established, would not by itself show how many participants read the exact D2 text, how many spoke for an affected organisation, how many were resource holders, or how many interests outside the room were exposed to the proposed rule.

The material available here does not disclose the denominator needed for those questions. It would be unsafe to infer who was represented, who remained silent, who objected or who consented. Silence has many possible meanings: agreement, uncertainty, lack of notice, language barriers, limited staff time, a judgement that speaking would not alter the result, or simple absence. The record does not select among them. Nor can the label “multi-stakeholder” transform attendance into a regional mandate.

Lu Heng’s controlling participation doctrine supplies the correct frame. Participation is evidence; it is not a transfer of sovereignty. A process can be open and still have an incomplete denominator. It can contain expert contributions and still lack authority over non-participants. It can achieve a valid internal consensus under its own rules without acquiring legislative capacity. This does not make participation worthless. It makes its evidential role precise.

The exact version is part of the denominator problem. Asking how many people “participated in the proposal” is not the same as asking how many reviewed the exact Draft 2 wording before the declaration. If the object shifts during a process, the affected-party review question must follow the final object. That is why publication time, version identity, review interval and preserved text belong in the same evidential chain. Without them, even a participant count could mislead.

The September boundary is the honest stopping point

The declaration is a natural narrative hinge, but it is the endpoint of this account. The article does not use later institutional developments to make the September judgement appear wise, futile, ominous or inevitable. Hindsight can corrupt the evidence in two directions. A later controversy may encourage readers to treat an earlier procedural choice as the first move in a settled plan. A later institutional success may encourage the opposite assumption, that earlier gaps were harmless. Neither inference belongs here.

Stopping at September preserves the causal order. First, there was a versioned draft. Then there was a procedural declaration about that draft. Questions about any subsequent process step, corporate act, contractual effect or operational consequence require their own records and authority tests. They cannot be projected backwards to enlarge or diminish what the September declaration meant at the time.

The discipline is especially important because the proposal concerned prerogatives. If a later event is allowed to colour the earlier text, the existence of practical power can be mistaken for proof of legal power. A registry may be able to affect operational dependencies because operators need accurate number-resource records and reliable services. That ability makes procedural safeguards more important. It does not make the ability sovereign, and it does not make every contemplated use of it valid.