Summary

  • On 12 October 2011, Resolution 201110.134 named four Board members and AFRINIC's Legal Counsel to review the company's bylaws and propose improvements, citing feedback at AFRINIC-14 without identifying the feedback or defining a detailed charter.
  • AFRINIC's later publications describe a genuine consultation sequence: a first draft, 63 comments, committee and Board review, another draft, a further comment window and a published 2012 text. Conflicting dates and missing proposal-level records prevent a clause-by-clause reconstruction.
  • The evidentiary gap matters because a committee is useful as a provenance bridge only when members can follow an issue from submission through evaluation, recommendation, redline, approval and commencement. Aggregate participation numbers cannot supply the missing links.
  • AFRINIC was, and remains in this analysis, a private recordkeeper and coordinator. Its bylaw process could organize corporate decision-making and affect service continuity, but it could not create sovereign, regulatory, police, punitive, confiscatory or adjudicatory power over operators or networks.

Five names, one sentence, and a large assignment

Resolution 201110.134 is unusually compact. It says the Board acted in line with feedback from the community during the AFRINIC-14 meeting in Dar es Salaam. It then appoints a committee made up of Board members Diop, Andriamampianina, Ngnoulaye and Akplogan, together with Legal Counsel Radhakissoon. Their task was to review AFRINIC's bylaws and propose improvements where needed.

That is the event. The conflict lies in what the appointment leaves unspecified. The resolution does not quote the meeting feedback, enumerate the alleged defects, link to submissions or say which bylaw provisions required attention. It supplies no chair, secretary, term, quorum, voting rule, conflict rule, public-member seat, acceptance record or stated deliverable. It does not explain how a concern would enter the committee, how the committee would classify it, what standard it would apply, how its recommendation would reach the Board, or what corporate approval would make a revised text operative.

The stakes are practical rather than sovereign. AFRINIC is a private company that performs narrow bookkeeping and coordination functions for Internet number-resource records. Its bylaws distribute decision rights within that company: among members, directors, meetings, proxies and committees. Weak documentation can make fees, records, service obligations and remedies harder to predict. For network operators, corporate disorder can spill into uncertainty around registry services, including resource records, RPKI and WHOIS or RDAP continuity.

None of that turns AFRINIC into a state or gives its committee power to regulate networks, punish users, confiscate assets or adjudicate public rights.

The core question is therefore narrower and more demanding than whether AFRINIC announced a review. Can an outsider connect the originating concern to a committee evaluation, the evaluation to recommended wording, the wording to a public draft, comments to reasoned dispositions, and the resulting text to an identifiable approval and commencement record? The surviving record supplies several parts of that chain. It does not supply the entire bridge.

What the October record establishes—and what it cannot establish

The current Board-resolution page and its archived snapshot establish AFRINIC's published account of Resolution 201110.134. They show the wording attributed to the Board and place the resolution within a meeting sequence. An adjacent resolution says the 12 October 2011 teleconference adjourned at 17:20 UTC. That helps locate the appointment in time.

Those pages do not function as a complete signed minute packet. They do not show the attendance list, quorum calculation, proposer, seconder, vote, recusals or conflicts for the committee appointment. They do not establish that every corporate requirement was satisfied. Official publication proves what AFRINIC recorded as having been said or done; it cannot, by its own status, prove legality, valid constitution, competence, representation, effectiveness or entitlement to public-law deference.

This distinction is especially important where institutional language sounds larger than the underlying act. A Board of a private company can appoint an advisory committee to examine that company's documents. It can assign work, receive advice and use corporate procedures to put proposed text before members. The label “Board resolution” does not elevate the appointment into legislation, and “community feedback” does not create a public mandate. The resolution organizes an internal review. It cannot enlarge AFRINIC's authority over operators, routed networks, address assets or the public.

Nor does the sparse appointment prove that the committee lacked fuller terms in private. A chair may have been chosen later. Working practices, legal advice, correspondence, draft notes or conflict arrangements may have existed without appearing on the public page. The responsible finding is not that these things never existed. It is that the publicly available chain reviewed here does not expose them, so an outsider cannot rely on them when reconstructing the process.

The five named people must also be described with care. The resolution calls Diop, Andriamampianina, Ngnoulaye and Akplogan Board members and identifies Radhakissoon as Legal Counsel. It does not allocate committee offices or specify whether counsel voted. It does not identify individual proposals, attendance, drafting authorship or personal positions. Any attempt to credit a later clause to one of these people would outrun the record.

The missing origin: feedback from AFRINIC-14

The resolution's opening premise is feedback during AFRINIC-14 in Dar es Salaam. The official meeting report places that event on 4–10 June 2011. It reports more than 200 participants from 51 countries, including 39 African countries, and records elections on 9 June. These figures establish the breadth AFRINIC chose to report about the meeting. They do not identify the bylaw concerns later invoked in October.

The report's searchable text does not set out the relevant feedback. That silence has two limits. First, it prevents a reader from determining whether the concerns involved elections, meetings, proxies, Board structure, member powers, committees, fees, drafting errors or something else. Second, it does not prove that no concern was voiced. A selective conference report may omit a discussion, while notes, recordings, correspondence or private records may once have captured it.

The difference between “not visible here” and “never happened” is fundamental. If the first link in a provenance chain is missing, later stages cannot retroactively specify it. A 2012 clause may address a real concern, but the clause alone cannot tell us whether that concern came from AFRINIC-14, from a committee member, from counsel, from the Board, from a later commenter or from another drafting source.

Attendance figures do not solve that attribution problem. More than 200 participants from 51 countries may indicate a geographically broad gathering, but attendance is not a vote. It is not a denominator for all African network operators, all AFRINIC members or the public. The source does not tell us which participants could submit bylaw feedback, how many did so, what standing they claimed or whether any position was tested against a member vote.

The word “community” can be useful shorthand for a loose field of participants. It cannot do analytical work that the record does not support. It cannot transform attendees, mailing-list contributors or commenters into a sovereign constituency. It cannot establish representation merely by being repeated. A serious account records who could submit, what was submitted, how it was assessed, who decided and what corporate approval followed.

The missing specificity matters at the outset because the appointment's legitimacy as a useful corporate procedure depends less on prestige than on traceability. If the October resolution had linked a list of concerns, an approved charter and a timetable, members could test whether the committee stayed within scope. Without those materials, the process must be reconstructed backward from later announcements, and backward reconstruction becomes less reliable with every missing handoff.

A committee should be a provenance bridge

A bylaw-review committee occupies a middle position. It receives problems or proposals, applies legal and operational judgment, and transmits recommendations into a corporate decision process. Its value is not that five names confer authority. Its value is that the committee can preserve reasons while turning diffuse concerns into precise text.

The ideal chain has ten distinguishable links. First comes an originating defect or proposal. Second comes an identified submitter and a clear account of that submitter's standing. Third is committee receipt and classification. Fourth is legal and operational analysis. Fifth is a committee recommendation. Sixth is a Board response. Seventh is a public draft accompanied by a redline. Eighth is a comment-and-response matrix. Ninth is a member-facing approval instrument. Tenth is the final text, commencement date and transition arrangement.

Not every link must disclose privileged legal advice. Confidential counsel can remain confidential while the non-privileged proposition, outcome and reason are recorded. Nor must every comma have an elaborate history. The demand for provenance should focus on material changes in rights, duties, control, meetings, voting, liability, fees, oversight or continuity.

For each such provision, a reader should be able to ask the same questions. What problem was identified? Who raised it? Did the committee accept, modify, reject or defer the proposal? Why? What wording changed? Did the Board agree? What did members receive? What threshold applied? When did the result begin to operate? Where the record does not answer one of these questions, the gap should be labelled rather than filled by institutional reputation.

Resolution 201110.134 publicly establishes an appointment, named membership and a broad task. The AFRINIC-14 report establishes a meeting but not the originating bylaw concerns. Later AFRINIC materials establish consultation milestones in aggregate. The 2007 and 2012 texts establish different endpoints. The principal weakness is in the connections among them: no visible issue register, proposal identifiers, committee minutes, disposition matrix, clean redline, final report or complete approval packet joins the stages clause by clause.

That finding does not erase the visible process. On the contrary, it makes the record more precise. There is enough evidence to say that AFRINIC announced a real sequence extending beyond the October appointment. There is not enough to assign every 2012 change to the five-person committee, to AFRINIC-14 feedback or to a particular public comment.

The first draft and the unresolved May–June conflict

AFRINIC's later published review timeline says a first draft was published on 7 May 2012. It describes a comment period from 8 May to 8 June and says 63 comments were received, ranging from editorial matters to structural issues. These are meaningful procedural claims. They show that the surviving official account is more substantial than a bare appointment followed by a finished document.

The timeline also describes a succession of tasks. The Review Committee and Board were to review and evaluate comments. A Board response would follow. Another draft and comment window would come next. A final review involving the Board and Review Committee would precede Board adoption and publication with a transition mechanism. Read as a process map, this assigns distinct stages to reviewers, directors, commenters and final publication.

There are two reasons not to treat the timeline as a complete contemporaneous ledger. The currently exposed page carries a 2026 last-modified marker, so it is evidence of AFRINIC's surviving account rather than a contemporaneously fixed timestamp for each 2012 event. More importantly, the page does not itself provide committee minutes, the first draft, an input-response matrix, the Board's detailed response, a redline, voting records or a final committee report within the bounded record reviewed here.

A dated announcement from 25 October 2012 complicates the first-draft chronology. That announcement says the process began with a first draft published in June 2012. June conflicts with the later timeline's specific date of 7 May and its comment period beginning on 8 May. The October message also says comments on the first draft remained available through a linked compilation, but that compilation is not available in the record reviewed for this analysis.

The conflict should not be smoothed into a single confident date. One possible explanation is loose retrospective wording in the announcement; another is a distinction between release, circulation and formal comment timing; another is an error on one of the pages. The evidence does not choose among them. The accurate chronology therefore carries both descriptions: 7 May, with comments from 8 May to 8 June, according to the later timeline; June, according to the 25 October 2012 announcement.

This may seem like a minor calendrical discrepancy, but dates are part of procedural provenance. They determine how long participants had to read a draft, which version comments addressed, and whether a response preceded or followed a revision. When the record does not reconcile its own milestone dates, claims about sequence should remain qualified.

Sixty-three comments are evidence of input, not a decision ledger

The reported 63 comments are the strongest numerical indication of participation in the first consultation period. The number suggests that the draft drew more than token attention and that commenters raised both editorial and structural matters. It supports the fair inference that a consultation took place and that the review had material to consider.

But 63 is an aggregate. It does not tell us whether 63 people each submitted once, whether some people submitted multiple comments, which participants were members, whether organizations submitted, or how the comments were distributed across provisions. It supplies neither the denominator of eligible or interested participants nor a measure of agreement. Above all, it does not reveal how each comment was handled.

For provenance, a comment becomes useful when it has a stable reference and a disposition. “Accept” should point to changed text. “Modify” should explain the adjustment. “Reject” should state a reason. “Defer” should identify the future route. Where several submissions raise the same issue, grouping may be sensible, but the grouping should remain visible. Without that structure, a count demonstrates volume, not influence.

This is why consultation cannot be converted into authority. Public comments can improve information, reveal drafting defects, expose operational consequences and make a private company's decisions more accountable. They do not make the commenters legislators. They do not confer public-law jurisdiction on AFRINIC. Nor do they authorize coercive powers over people who did not participate.

The relevant standard is not maximal disclosure. Some comments may contain personal details; some legal exchanges may be privileged; some editorial suggestions may be too trivial to catalogue at length. A well-designed register can protect legitimate confidentiality while showing the substance of material proposals and their outcomes. The absence of such a visible register here limits reconstruction without proving that reviewers ignored the submissions.

The October announcement's reference to a linked compilation is therefore important but not conclusive. It indicates that AFRINIC said comments had been made available. Because that compilation is not available in the bounded record, this analysis cannot use it to identify authors, content or dispositions. The correct treatment is to record the reference and the present evidentiary limit, not to infer either transparency or suppression beyond what can be shown.

A second draft and another shifting timetable

The 25 October 2012 announcement says a new draft had been published and opens a 15-day comment period ending on 9 November. It then assigns review by a bylaw-review working group to 12–19 November, followed by final review and Board ratification during 20–25 November. This is a relatively concrete forward schedule. It indicates a further opportunity for comment and names planned handoffs after the comment window.

Yet the announcement's revised timetable differs from the earlier schedule retained on AFRINIC's later timeline page. The bounded records do not reconcile the dates or prove that each planned milestone occurred precisely as announced. A timetable is evidence of intended procedure; it is not, without completion records, proof of performance.

The terminology also shifts. The October resolution names a committee. The later announcement refers to a bylaw-review working group. The timeline says the Board was considering extending the Review Committee for the remainder of the process to people who had expressed interest in Serekunda. That may indicate continuity, enlargement, informality or a change in label. It does not prove who was added, when an addition occurred, which appointment instrument governed, or whether any new participant could vote.

This matters because committee composition is part of the control surface. A recommendation from the five named appointees is one thing; a recommendation from a larger or differently constituted group is another. Expansion may improve expertise and participation, but it should be recorded. Without an appointment record, terms and status, an outsider cannot tell whose judgment a later draft embodies.

The time allowed for the final comment window also deserves exact treatment. The announcement describes 15 days and names 9 November as the endpoint. It does not establish how many submissions arrived, how they differed from the first set, or which revisions followed. Nor does it prove that Board ratification occurred between 20 and 25 November. The schedule should be reported as an announced plan, not silently converted into a completed chronology.

Taken together, the May/June conflict and the revised October/November schedule show why procedural history needs versioned records. Each draft should carry a publication time, stable identifier and change log. Each consultation should attach to one identified version. Each announced stage should later receive a completion record. Otherwise, even good-faith updates can leave future readers unsure which calendar reflects the actual path.

Different 2007 and 2012 texts do not prove who caused each change

The archived 2007 and 2012 bylaws are plainly distinct published corporate texts. The 2012 version contains more explicit structures, including Nomination and Election Committees, expanded member-meeting and proxy provisions, a nine-director structure, a Council of Elders and language permitting Board committee appointments. These differences establish that the later document did more than reproduce the earlier one.

They do not establish provenance. A before-and-after comparison can identify inserted, deleted or revised text. It cannot reveal, by itself, which person proposed a clause, which concern it answered, whether counsel drafted it, whether the committee endorsed it, whether the Board altered it, which public comment affected it or what approval made it operative.

The distinction is easy to miss because endpoints are tangible. A reader can place two texts side by side and see change. The middle of the process—evaluation, reason, authorship and approval—is less visible. But those middle links are where accountability lives. If a provision changes member voting, proxy use, committee power or Board structure, the reason and approving constituency matter as much as the final wording.

Article 7.6(vi) of the 2012 text is particularly relevant. It assigns members the power to consider and approve by Special Resolution, if appropriate, proposals to revoke, amend or replace the Constitution. The same 2012 text defines a Special Resolution as approval by at least 75% of votes of members entitled and voting on the matter. These provisions describe a member-facing approval architecture.

The bounded record does not include the specific Special Resolution, complete notice, quorum record, voting denominator, tally or commencement record for the final 2012 instrument. It would therefore be wrong to state that a particular 75% vote approved this version. The text tells us what the published rule says; it does not provide the event record needed to prove that the rule was followed in this instance.

Article 15.3(x) of the 2012 text says Directors may appoint committees for reasons and with terms of reference they consider necessary or desirable. That later provision cannot be used retroactively to prove the exact legal basis or terms applied to the October 2011 appointment. It is evidence of the later text's allocation of committee power, not a missing 2011 charter.

Nor can the differences between the two texts all be attributed to Resolution 201110.134. The resolution initiated a review and named five people, but the later process involved consultation, the Board, a possible expanded group and further drafting. Some changes may have arisen from sources not visible here. Attribution requires records, not proximity in time.

Approval, publication and effect must remain separate

Corporate processes often compress several acts into one narrative: a draft is reviewed, adopted and published, therefore it became the governing text. For accountability, those verbs should remain separate. Review means someone examined a proposal. Adoption means an identified body made a decision under a stated rule. Member approval, where required, means a constituency received notice and met a threshold. Publication makes text available. Commencement determines when it begins to operate. Transition explains how existing offices, decisions or procedures move into the new arrangement.

AFRINIC's later timeline describes Board adoption and publication with a transition mechanism. The October announcement forecasts final review and Board ratification. The 2012 bylaws describe member power over amendment or replacement through Special Resolution. These records show that approval and transition were recognized concepts. They do not, in the bounded materials, assemble the exact final packet for the 2012 change.

That missing packet would ideally contain the final recommended text, a clean redline against the previous version, the Board decision, member notice, the exact resolution put to members, the voting eligibility rule, quorum, tally, commencement date and a transition table. Such a packet would let a reader distinguish recommendation from authorization and authorization from publication.

The absence of that packet from the record reviewed here is an evidentiary limitation, not a verdict that the final text was invalid. Records may exist elsewhere or may later be produced through a defined inspection, dispute or litigation process. If they appear, the historical uncertainty should narrow. Until then, the disciplined conclusion is that the exact approval chain is not independently reconstructable from these public materials.

This discipline protects both criticism and fairness. It prevents an official publication from certifying its own legitimacy. It also prevents missing documents from being treated as proof of misconduct. The same standard applies in both directions: identify the instrument, actor, authority, threshold and validation record; do not replace any of them with an institutional label.

Private coordination, not public authority

The most serious category error would be to read a corporate bylaw process as a source of sovereign power. AFRINIC keeps and coordinates records associated with Internet number resources. That function can be important, technically sensitive and economically consequential. Importance does not make the company a legislature or regulator.

Resolution 201110.134 could organize advice about AFRINIC's own governance. A later corporate instrument could allocate responsibilities among the company, its members, directors and committees within applicable private-law limits. Neither act could create police power, prosecutorial authority, a right to punish, power to confiscate address assets, or jurisdiction to adjudicate public rights. AFRINIC does not become the sovereign of the networks whose records it coordinates.

The same boundary applies to consultation. A broad invitation to comment may improve a decision. A meeting across 51 countries may supply diverse knowledge. Sixty-three comments may reveal defects that a small committee would miss. These are informational and accountability benefits. They do not manufacture sovereignty. No volume of mailing-list participation converts a private company into a state.

This is why “community support” cannot cure a missing corporate instrument. If members must approve a change under a defined threshold, comments are not a substitute for that vote. If a committee requires terms, popularity is not a substitute for terms. If a service action affects a contractual or property interest, a slogan about consensus is not adjudication. Each kind of act must be tested through the authority and procedure actually applicable to it.

Keeping the boundary narrow also clarifies the remedy. Opacity does not justify giving the recordkeeper broader discretionary power. It argues for auditable, replaceable coordination: precise records, narrow terms, separable decisions, effective member remedies and continuity arrangements that do not depend on institutional mythology.

Why operators should care about a corporate paper trail

The bylaw review may look remote from routing operations. It is not the source of routing authority, and it does not govern the Internet. Yet corporate arrangements can influence who makes registry decisions, how members challenge those decisions, how fees are set, how meetings occur and how continuity is managed.

Uncertainty in those arrangements creates transaction costs. A member assessing a disputed decision must determine which body acted, under which version of the rules, after what notice and with what remedy. An operator assessing continuity risk must understand who can authorize expenditures, maintain records, appoint responsible committees or manage a transition. Lawyers, insurers, counterparties and technical teams all spend more when the decision trail is incomplete.

The risk reaches technical services through the operating institution, not through sovereignty. If corporate conflict disrupts staffing, contracting, access controls or decision-making, registry data and related services can become less predictable. RPKI, WHOIS and RDAP depend on accurate administration and resilient systems. AFRINIC's private status does not make operational failure harmless; it makes precise contractual and corporate accountability more important.

A reproducible provenance chain lowers the cost of trust. Members need not accept a committee's prestige if they can inspect the issue, reason, text and approval. Operators need not treat every bylaw dispute as a threat if commencement and transition are clear. The institution becomes easier to audit and, where necessary, to correct or replace.

By contrast, an opaque bridge concentrates interpretive power. When no proposal matrix or redline survives, current officeholders can more easily assert what a clause was meant to accomplish. When no exact approval record is visible, participants may argue from competing recollections. The resulting uncertainty is itself an operational burden even if no service interruption occurs.

The economic point is therefore modest but real: governance-document clarity is infrastructure-risk control. It does not transform corporate rules into public law. It reduces ambiguity around the private organization that maintains important records and services.

The strongest case for the process

A fair assessment should acknowledge that a volunteer or private-company review does not need to publish every internal exchange. Committees often work through informal drafts. Legal advice may be privileged. Detailed attribution can discourage candid discussion. A Board may reasonably appoint people with institutional knowledge, invite comment, revise text and present a finished instrument without turning every deliberation into a public transcript.

On that view, the surviving record is meaningful. The Board named responsible people. AFRINIC later described a consultation window and counted 63 comments. The account assigns evaluation to the Review Committee and Board, anticipates responses and further drafting, and a dated announcement opens another comment period. A distinct 2012 text was published. These elements support an inference that a genuine review process occurred.

That is a strong concession, not a rhetorical aside. The evidence is inconsistent with the claim that nothing happened between appointment and publication. It would also be unfair to infer that confidential legal work did not exist merely because it is not public.

The answer is proportionality. The minimum public chain need not expose privilege or every drafting conversation. It should expose material propositions and decisions. For a significant clause, members should see the originating issue, the proposed treatment, the committee's disposition, a reason, the resulting change, the approval route and the transition consequence. A summary count of comments cannot perform those functions.

The stronger the change, the stronger the need for traceability. Editorial corrections may warrant a short change log. Structural changes to member meetings, proxies, Board composition, nomination processes or committee powers call for more. A practical governance process can remain efficient while publishing stable proposal identifiers, concise reasons and redlines.

Thus the critique is not that AFRINIC had to expose every thought. It is that the surviving record does not provide enough non-privileged structure to trace important changes clause by clause. Aggregate consultation and final text show activity and an outcome. They do not, alone, prove proposal-level provenance or member authorization.

What a reconstructable process would have looked like

At appointment, the Board could have attached three short documents. The first would list the bylaw concerns attributed to AFRINIC-14, with neutral identifiers and enough context to understand each issue. The second would state the committee's term, membership, chair, quorum, conflict rules and whether counsel had a vote. The third would set a timetable and identify the required outputs: issue register, recommendations, draft, redline and final report.

During consultation, every material submission could receive a stable reference. The public register need not reveal protected personal information. It could show the provision affected, the substance of the proposal, submitter category, date and status. The committee's later matrix could mark each item accepted, modified, rejected, merged or deferred, accompanied by a concise reason.

Draft publication could then preserve version control. Each draft would state when it was released and which prior version it replaced. A redline would distinguish editorial cleanup from structural changes. Machine-readable mapping could connect each alteration to one or more accepted proposals, while acknowledging where the committee or Board initiated a change independently.

The Board's response would form another separate record. It would identify which recommendations were accepted, changed or rejected and why. If committee membership expanded after interest expressed in Serekunda, an appointment record would name the new participants, their status and the date on which they joined.

The final member packet would identify the exact text, notice, eligibility rules and threshold. If a Special Resolution applied, it would state the denominator and tally. The final publication would include commencement and a transition table, showing how existing offices, committees, notices and pending matters were treated.

Under that counterfactual, the reported 63 comments would become more than a headline number. A reader could follow a concern from submission to result. The May/June discrepancy could be resolved by a version log. The later timetable could be checked against completion notices. The difference between committee advice, Board choice and member approval would remain visible.

These measures are not administrative ornament. They make accountability portable. A future member, court, auditor or operator would not have to trust the memory of participants. The process could be tested from the record even after personnel changed.

The bounded unknowns

Several questions remain open. The first is the content of the AFRINIC-14 feedback invoked by Resolution 201110.134. The second is the committee's contemporaneous charter: its leadership, duration, quorum, conflict arrangements, voting practice and consultation mandate. The third is whether its membership changed, including any extension to people who expressed interest in Serekunda.

The full content and authorship of the first 63 comments are not established here, nor are the later comments. No proposal-by-proposal response, detailed Board response, clean redline or final committee report is available in the bounded public chain. The May/June first-draft conflict remains unresolved, as does the relation between the earlier schedule and the revised October/November timetable.

At the approval end, the specific member resolution, notice, quorum, voting denominator, tally, commencement record and transition details are not present. As a result, no particular 2012 clause can safely be attributed to a named committee member or individual comment.

Each unknown should remain exactly that: unknown. It should not be converted into an accusation, a presumption of validity or an invented causal link. Records discovered later could fill some gaps. The historical account should be capable of revision when better evidence appears.

The limits also define what this analysis does not own. It is not a general comparison of every provision in the 2007 and 2012 bylaws. It is not a complete election-process history. It does not extend into later institutional crises or later review initiatives. Its subject is the provenance bridge created in October 2011 and the review process leading into the 2012 instrument.

A real sequence, an incomplete bridge

The surviving public account contains enough pieces to reject two extremes. It is not merely a one-sentence appointment followed by silence: AFRINIC later published a timeline, reported 63 comments, announced another draft and review window, and published a distinct 2012 bylaw text. But it is also not a fully reproducible record of how each concern became each clause.

Resolution 201110.134 shows the beginning in broad outline. The review timeline and October announcement show middle stages in aggregate. The two bylaw texts show different endpoints. What remains weak is the joining evidence: the originating issue register, committee charter and minutes, proposal dispositions, Board response, redline and exact approval record.

That makes the committee visible as a procedural bridge but leaves much of its decking uninspected. The conclusion is evidentiary, not metaphysical. It does not say the five appointees did no work. It does not say private records never existed. It does not say every 2012 change was defective or invalid. It says an outsider cannot independently reconstruct the full proposal-to-clause chain from the surviving public record reviewed here.

For a private recordkeeper and coordinator, that is the right level of scrutiny. AFRINIC's work can matter greatly to operators without acquiring sovereign character. Corporate consultation can improve decisions without creating a demos. Official pages can prove publication and recorded acts without certifying their own legitimacy. The durable remedy is a narrower, more auditable process in which every material handoff leaves a visible trace.