AFRINIC's Bylaws Before and After Crisis

Summary

• AFRINIC published constitutions in 2007, 2012, 2016 and 2020, but the operative public text remains the version that took effect before the litigation, receivership and repeated difficulty restoring an elected board.
• The 2020 amendments were substantive: fourteen special resolutions obtained the stated 75 percent majority and dealt with matters including proxies, the Nomination Committee, member powers, directors and document authentication. They should be credited, but not mistaken for a response to failures that had not yet unfolded.
• The crisis tested a combined system of company constitution and Mauritius law. Courts and an Official Receiver became necessary when internal organs could not reliably restore ordinary governance.
• AFRINIC's 2026 review created a timetable for comments, a draft, a final report and a proposed special meeting. As at 11 July 2026, it remained unfinished. A consultation becomes constitutional repair only after members receive a sufficiently precise text, lawfully adopt it, register it where required and can use its remedies in practice.

The word "after" needs an immediate qualification

The title of this article contains a trap. It invites a neat comparison between an old AFRINIC constitution and a new constitution enacted after the organisation's governance crisis. The public record does not support that comparison. AFRINIC's published constitution page identifies a fifth version effective in December 2020, replacing the December 2016 version, and links earlier editions. The major litigation and receivership stress that followed was not answered by a completed post-crisis constitution before the evidence cutoff for this analysis.

There is a post-crisis review. AFRINIC's April 2026 consultation notice set out a sequence running from initial comments to a draft, public review, a final report, notice and a planned special general members' meeting. That sequence matters. It shows an institution attempting to convert painful experience into constitutional revision. But a timetable is not a constitutional instrument, a committee is not a member resolution, and a draft is not an operative limit on power.

The honest before-and-after analysis is therefore asymmetric. On the "before" side are enacted texts from 2007, 2012, 2016 and 2020. In the middle is a crisis that tested whether the 2020 settlement could produce a board, elections, accountable control and continuity under severe conflict. On the "after" side is an unfinished reform process whose final language, voting denominator, adoption record and registration were not yet proved as at 11 July 2026.

That asymmetry is not a technical footnote. It is the central governance finding. Institutions often describe the start of review as evidence that a crisis has been absorbed. It has not. Review creates an opportunity to repair. Only an enacted rule changes who may act, on what conditions, with which notice, subject to which remedy and under whose supervision.

The distinction also protects reformers from an unfair claim. They should not be judged as if a text that was still under discussion had already failed. But neither should the institution receive credit for constraints that members had not adopted. The correct test is staged: what the old text said, what the crisis exposed, what the draft proposes, what members approve, what the company registers, and what later conduct proves.

Four versions show amendment capacity, not continuous adequacy

AFRINIC did not operate under a constitution frozen at incorporation. Its archive shows repeated constitutional change. The 2012 bylaws provide an early baseline for the company's entities, membership structure, definition of the Registration Service Agreement, member powers, board appointment and removal, and winding up. The 2016 bylaws provide the immediate predecessor to the 2020 settlement. The current page says the 2020 text was the fifth constitution.

This history rebuts the claim that AFRINIC was constitutionally inert. Members and corporate organs had a route for amendment, and they used it. That matters because a constitution incapable of change would make every later failure seem inevitable. AFRINIC's record instead presents a more demanding question: why did repeated amendment not produce a settlement resilient enough for the crisis that followed?

One answer may be that no constitution can anticipate every adversarial event. That is true. Corporate rules are not disaster scripts. They allocate ordinary powers, define membership, set meeting and election rules, and supply procedures for exceptional decisions. Host law fills gaps and courts resolve disputes. A text should not be condemned merely because litigation occurred.

Another answer, however, is that amendment quality cannot be measured by version count. A new edition may clarify drafting without changing incentives. It may solve yesterday's procedural inconvenience while leaving control concentrated. It may improve formal voting while leaving remedies slow. It may give an organ power without specifying what happens when that organ is absent, conflicted or unable to act. It may create a nomination process without protecting an election from later contest over eligibility, ballots or implementation.

Version history is therefore evidence of constitutional activity, not proof of constitutional adequacy. The relevant audit is functional. For every change, ask which failure it was meant to prevent; who gained or lost discretion; whether the change reduced ambiguity; whether affected members could challenge misuse; and whether the rule still worked when ordinary trust disappeared.

This is why authenticated redlines and implementation records matter. A reader needs more than four PDFs. The strongest record would connect each altered clause to the resolution that approved it, the meeting notice, explanatory memorandum, votes for and against, abstentions, proxies, effective date, registry filing and later cases in which the clause was applied. Without that chain, comparison can identify text but cannot fully explain institutional effect.

The 2020 settlement deserves precise credit

The current AFRINIC page states that fourteen special resolutions obtained the required 75 percent majority at an online annual general members' meeting on 18 September 2020. The organisation says those resolutions produced the fifth constitution effective that December. The topics included proxies, the Nomination Committee, member powers, director appointment and other corporate mechanics, as well as authentication of documents.

Those facts matter for two reasons. First, the 2020 constitution was not a cosmetic republication. It resulted from formal resolutions on governance subjects that later became central to AFRINIC's ability to function. Second, the record establishes a high approval threshold for those resolutions. It is wrong to say AFRINIC entered crisis without any recent attempt at constitutional maintenance.

But precision cuts both ways. A 75 percent threshold is a proportion of votes counted under the applicable rules, not necessarily 75 percent of all Resource Members. The public summary does not, by itself, provide every denominator needed to evaluate participation: eligible members, valid ballots, abstentions, excluded voters and proxy use. Supermajority language can show that the legal threshold was met while leaving the breadth of practical participation uncertain.

The date also fixes the direction of causation. September and December 2020 precede the principal 2021-2025 crisis. The amendments cannot honestly be described as lessons enacted after that crisis. They were the rules the crisis tested. If a provision later proved ambiguous or difficult to operate, that is evidence about the 2020 design. It is not evidence that the 2020 drafters knowingly ignored a later court sequence.

This makes the 2020 text more valuable as an institutional experiment. It lets members ask whether recent reforms behaved as expected. Did revised proxy rules make member decisions more trustworthy? Did Nomination Committee changes produce a robust candidate process? Did director provisions handle vacancies and contested elections? Did member-power provisions make oversight usable during institutional paralysis? Did authentication clauses reduce or merely relocate disputes about valid corporate acts?

The answer cannot be inferred from headings alone. It requires a failure-mode matrix, linking each crisis event to the exact clause and statutory provision that operated, failed, conflicted or required judicial supplementation. Such a matrix would be more useful than a broad claim that AFRINIC had "outdated bylaws." The problem was not simply age. The current text was months old when the crisis intensified. The problem was whether its allocation of authority and remedies remained executable under pressure.

A constitution is part of a legal system, not a private universe

AFRINIC is a Mauritian company limited by guarantee. Its constitution does not float above the law of its place of incorporation. The official consolidated Mauritius Companies Act 2001 supplies mandatory rules and legal powers concerning constitutions, directors, meetings, receivers and winding up. It can supplement the corporate text and, where mandatory law requires, override it.

This matters because constitutional commentary often treats a company's bylaws as if they were a complete political constitution. They are not. They are a corporate instrument inside a domestic legal order. They bind and empower according to company law, the company's legal status, valid member decisions and applicable court orders. Their meaning in a disputed case cannot be separated from statute.

The host-law layer does not diminish the importance of AFRINIC's constitution. It makes the constitution real. A clause on appointment matters because the law recognises corporate office. A member resolution matters because statute and constitution define how members act. A receiver matters because a court acting under law can place authority in a person other than the ordinary board. Registration matters because corporate change is not established only by a website update.

It also identifies the limit of institutional rhetoric. Neither continental service scope nor technical importance displaces Mauritian corporate law. AFRINIC may perform a regionally important registry function, but its legal organs remain those of a company. Participation by stakeholders can provide expertise and legitimacy, yet it does not create a new source of power outside the instruments that actually authorise corporate action.

The statute should not be used as a shortcut either. Pointing to a general court power does not prove that every order was correctly made or implemented. Pointing to receiver provisions does not resolve election eligibility. Pointing to meeting provisions does not establish the validity of a particular ballot. Each event must be tied to the law and order applicable on its date.

That discipline is especially important in a long dispute. Company status, officeholders, interim orders and appeal outcomes can change. A proposition correct in September 2023 may be altered by an order in October 2024. A planned election may be postponed or contested. The legal record must therefore be dated. "The court decided" is not enough; the analyst must say which court, on which date, in which procedural posture and with what operative consequence.

The crisis tested whether internal authority could regenerate itself

The most important constitutional function under stress is regeneration. A company can survive disagreement if it still has a lawful way to identify members, nominate candidates, conduct a vote, determine the result, seat directors and resolve challenges without destroying ordinary operations. When that chain breaks, every later act becomes harder because the organ needed to authorise it is itself disputed.

AFRINIC's crisis exposed precisely this problem. The issue was not only whether one board decision was right. It was whether the institution could restore an ordinary governing board through procedures accepted as valid by members, officers and courts. Once receivership and election litigation became central, the constitution's vacancy, appointment, meeting and electoral provisions were no longer routine housekeeping. They were the mechanism by which corporate authority was supposed to become normal again.

An AFRINIC notice dated 16 October 2024 reported that the Supreme Court had set aside AFRINIC's appeal, restoring the judgment of 12 September 2023 and the appointment of the Official Receiver with a two-month mandate to organise board elections under AFRINIC's constitution. The notice is useful because it records the immediate institutional understanding of the result and the restored election mandate.

It is not the full judgment. That limitation is important. An institutional announcement may accurately state the order's effect, but it cannot replace the court's reasoning, conditions, subsequent variations or complete docket. A rigorous constitutional history needs the judgments and orders themselves, not only communiques issued by parties affected by them.

Even with that limitation, the event reveals a decisive fact: internal processes alone had not restored ordinary governance. A court-appointed officer was required to conduct elections under the constitution. Host law did not erase the bylaws; it provided an external actor to make their electoral mechanism executable.

That can be read as evidence of resilience. Mauritius law supplied a lawful route when the company was stuck. The registry function did not have to depend entirely on self-help by a contested board. Courts remained available, and an Official Receiver could be assigned a specific task.

It can also be read as evidence of constitutional fragility. If the company repeatedly needs judicial direction to identify who may organise the vote, who may participate, or how results take effect, then the internal rule set has not contained the dispute. Both readings can be true: external law prevented total paralysis, while reliance on external intervention showed that the constitution did not regenerate governance by itself.

Elections require more than a calendar

A constitutional election is a chain of verifiable acts. The calendar is only the visible surface. Beneath it sit the voter roll, eligibility rules, nomination authority, candidate checks, notice, ballot custody, proxy rules, counting, challenges, certification and assumption of office. A failure in any link can prevent the board from becoming an accepted corporate fact.

The AFRINIC record should therefore resist a recurring simplification: that scheduling an election equals restoring member control. An election can be ordered and still fail to produce a durable board. It can produce ballots without producing accepted authority. It can meet a formal date while leaving disputes over exclusions, proxies or certification unresolved.

The constitutional design question is not whether every dispute can be eliminated. It is whether disputes can be isolated and decided without invalidating the whole institution. A strong system distinguishes correctable errors from fatal ones. It sets deadlines for challenges. It identifies an independent decision-maker. It preserves evidence. It states whether uncontested seats can take effect while another seat is challenged. It limits the ability of any interested organ to alter the voter roll or candidate field.

These safeguards matter more in an association whose service decisions affect infrastructure beyond those who attend meetings. Members possess corporate rights, but operators, customers and third parties may rely on registry continuity. Election litigation should not become a switch through which every technical service or public record is put at risk.

That does not mean insulating AFRINIC from courts. It means designing corporate remedies so that legal review and operational continuity can coexist. A court should be able to correct an invalid election without creating uncertainty over every database entry. A receiver should be able to conduct a ballot without being forced to improvise corporate rules. Members should be able to challenge exclusion without waiting until the resulting board has made months of contested decisions.

The public record available here does not supply the denominators needed to assess this chain. It does not provide a complete list of eligible Resource Members for each attempted election, ballots issued and received, abstentions, proxies, rejected votes, challenged candidates, reasons for exclusion or final implementation. Those are not marginal statistics. They are the evidence by which a member-governed company demonstrates that its principal organ can be renewed.

Membership is a legal relationship, not an atmospheric community

AFRINIC's constitution distinguishes corporate membership and connects it to the Registration Service Agreement. That architecture matters because the word "community" is often used too broadly. The technical community can discuss policy and provide knowledge. Stakeholders can comment. The wider public can be affected. But corporate votes belong to legally defined members under the applicable constitution.

This boundary is not anti-participatory. It prevents participation from being misdescribed as authorisation. A consultation can improve a draft. It cannot substitute for the member act required to amend the constitution. A committee can synthesise views. It cannot claim the same legal standing as a properly convened meeting. A large number of comments can show attention. It cannot answer whether the required voting threshold was reached.

The boundary also protects absent members. If a small, active group dominates discussion, their expertise should be heard without treating attendance as a mandate from every holder or network in the service region. The constitution should identify which decisions active entities may influence, which decisions members must approve, and which operational duties belong to staff or directors.

At the same time, formal membership is not sufficient evidence of accountability. A right that is too costly, slow or opaque to use may exist only on paper. Members need timely notice, intelligible proposals, access to the controlling text, an accurate roll, usable voting methods and a remedy if an organ refuses to implement a valid result.

This is where the 2020 denominator gap becomes relevant. Fourteen resolutions reaching a 75 percent threshold establishes formal success under the announced procedure. To evaluate institutional legitimacy more fully, AFRINIC should also publish the eligible electorate, participation rate, proxy use, abstentions and excluded ballots for each resolution. Those figures would not displace the legal threshold. They would reveal how broad the practical consent was.

Post-crisis reform should make that distinction explicit. Public consultation measures input. Constitutional adoption measures member approval. Registration establishes the operative corporate text. Later use measures whether the reform works. Combining all four into a generic claim of "community endorsement" would repeat the ambiguity the reform is meant to cure.

The 2026 review is meaningful precisely because it is unfinished

The 2026 consultation notice should be read neither cynically nor triumphantly. It is meaningful evidence of a repair effort. It announces a review committee, comment periods and a route toward a special meeting. It gives members and observers dates against which progress can be checked. After years in which governance recovery was contested, an organised constitutional review is a necessary step.

Its incompleteness is equally meaningful. The notice was published on 18 April 2026. Initial comments were extended to 17 May. Public comment on a draft was scheduled from 28 June to 17 July, followed by later stages and a planned September special meeting. On 11 July, the public-comment window had not even reached its scheduled end. No final member-adopted constitution could therefore be inferred from the timetable alone.

Future reporting must preserve that temporal discipline. If a draft appeared on 28 June, that would prove publication, not adoption. If a committee delivered a final report, that would prove recommendation, not approval. If meeting notice was issued, that would prove convocation, not a valid vote. If members voted, that would still require verification of threshold, implementation and any required filing.

The process can gain legitimacy by publishing more than a clean final document. It should publish an authenticated redline from the 2020 constitution, a clause-by-clause rationale, the crisis failure each change addresses, comments received, the committee's response, minority positions, legal review and the exact voting effect of every resolution.

Such a record would reduce the risk of reform by narrative. Members could see whether a proposed safeguard actually changes operative words or merely adds aspirational language. They could identify provisions that shift power toward the board, receiver, members, committee or staff. They could test whether new emergency authority has objective triggers and an end date.

The unfinished status also creates room for better drafting. Once a draft is adopted, ambiguity becomes expensive. During consultation it can still be corrected. Treating the process as completed too early would weaken that opportunity by converting questions into perceived attacks on a settled outcome.

What genuine post-crisis repair would contain

A serious post-crisis constitution should begin with continuity of lawful corporate action. It should define what happens when the board lacks quorum, all or most seats are vacant, appointments are disputed, an election is set aside, or a court places the company under an external officer. Emergency rules should be narrow, time-limited and tied to restoration of ordinary governance.

It should then protect election integrity. The voter roll should have a fixed qualification date, transparent categories and a correction route. Nomination criteria should be known before candidates apply. Disqualifications should be reasoned and reviewable. Ballot custody and counting should be independently auditable. Proxy rules should be clear. Results should identify the denominator, and challenges should have short, realistic deadlines.

The constitution should separate corporate legitimacy from technical continuity. A contested board does not make every registry record false. An election challenge should not interrupt RDAP, WHOIS, reverse DNS or RPKI. Staff authority to preserve routine services during a vacancy should be explicit and limited, while material policy, enforcement or asset decisions should face stronger controls.

It should also separate recordkeeping from punishment. AFRINIC needs authority to maintain accurate registrations, correct fraud, administer agreements and comply with lawful orders. But high-consequence action against live resources should be grounded in a specific contract or policy condition, accompanied by notice where possible, and subject to independent review. Broad company entities should not operate as a reserve sanction.

Member remedies need practical force. It is not enough to say members control the company if they cannot obtain records, call a meeting, correct an electorate, challenge a conflicted decision or enforce a valid result. The constitution should identify forum, standing, deadlines and interim protection.

Conflicts of interest need treatment proportionate to the stakes. Directors, candidates, committee members, receivers and advisers may have relationships that do not automatically disqualify them, but disclosure and recusal rules should be explicit. A crisis produces suspicion; transparent rules are the way to prevent every disagreement from becoming a legitimacy dispute.

Finally, amendment itself needs an evidence trail. Every future constitutional version should have a stable date, resolution list, voting record, authenticated text and filing reference. AFRINIC's archive is already a useful foundation. The post-crisis settlement should make version-level accountability routine rather than reconstructing it after the next dispute.

Reform can redistribute power while claiming to constrain it

Crises create pressure for decisive authority. That pressure can improve institutions when it produces clear succession and review. It can also concentrate power. A new clause may be described as a safeguard because it lets someone act quickly during paralysis, while its practical effect is to move decisions away from members or ordinary courts.

Every proposed reform should therefore be tested for both limitation and redistribution. Who may trigger the clause? Who decides that the trigger exists? What evidence is required? How long does emergency authority last? Which decisions remain prohibited? Who reviews use? Can the beneficiary of the power extend it?

This test is more reliable than asking whether a clause sounds modern. Governance documents often contain reassuring nouns: integrity, stability, accountability, continuity and community. The operative verbs determine the actual settlement. A clause that says an organ "may determine" a question has a different effect from one saying it "shall apply" published criteria. A clause allowing removal "for cause" needs a definition and a hearing route. A clause authorising interim appointments needs an election deadline.

The same scrutiny should apply to member power. More voting is not automatically more accountable if the electorate is unclear, notice is weak or a small active group can bind a much larger class without transparent participation data. Conversely, requiring every operational act to go to a member vote would make the registry unmanageable. The aim is not maximum democracy at every layer. It is authority matched to function, impact and remedy.

The post-crisis constitution should also avoid converting technical continuity into institutional immunity. AFRINIC performs important services, but the importance of those services is a reason for replication, auditable custody and succession planning. It is not a reason to place directors or corporate decisions beyond ordinary legal accountability.

A resilient institution is one whose function can survive a failed board, a disputed election and judicial correction. If continuity depends on treating the incumbent organ as untouchable, the constitution has protected the gatekeeper rather than the registry.

Courts are a safeguard, but not a complete governance model

Mauritius courts and the Official Receiver supplied an external route when AFRINIC's internal governance was unable to normalise itself. That should be recognised as a safeguard. Private institutions of regional importance remain subject to law. Members and affected parties are not limited to appeals inside the institution's own culture.

But repeated court management is expensive, slow and fact-specific. Judges decide disputes brought before them under available law. They do not ordinarily design the company's next nomination cycle, publish participation denominators or operate technical services. A receiver can carry out an assigned mandate, but receivership is not a permanent substitute for a board accepted under the constitution.

The right reform objective is therefore not to exclude courts. It is to reduce the number of questions that require a court simply to make ordinary governance possible. Clear vacancy rules, auditable elections, narrow emergency authority and effective internal remedies can contain disputes before they become existential.

External review remains essential for high-impact cases, conflicts that internal bodies cannot decide impartially, and questions of company law. Internal procedures should preserve evidence and provide interim stability so that court review corrects a bounded dispute rather than deciding the organisation's entire operating reality.

The official 2024 notice demonstrates both sides. The restored receiver mandate provided a lawful way forward. The need to restore it through Supreme Court proceedings showed how far ordinary governance had moved into litigation. A post-crisis constitution succeeds if future courts can interpret and enforce clear rules. It fails if it tries to make judicial scrutiny impossible or leaves judges to construct missing procedures from general language.

The missing evidence should shape the final text

Several important records were not available for this analysis. The first is a complete authenticated redline from the 2007 constitution through 2012, 2016 and 2020, paired with resolutions and implementation dates. The second is the full set of reasoned judgments and orders from 2023 through 2026. The third is the actual 2026 draft, comment ledger and response matrix as they stood through the full consultation period.

The fourth is electoral data: eligible members, ballot distribution, participation, abstentions, proxies, exclusions, challenges and certification for each attempted restoration. The fifth is evidence of whether later scheduled milestones occurred, including final resolutions, vote records, executed text and filing.

These absences limit conclusions. They do not justify saying nothing. They justify narrower findings. The 2020 constitution is the last proved operative public version. It predates the central crisis. Courts and a receiver were required in the effort to restore elections. A 2026 review was active but unfinished as at 11 July. Beyond those points, claims should remain conditional until the relevant instruments are produced.

The missing material should also inform institutional design. If future researchers cannot identify which clause governed an election, the version system is inadequate. If members cannot reconstruct who was entitled to vote, the electorate system is inadequate. If the public cannot tell whether a consultation comment changed the draft, the review record is inadequate.

Transparency here is not performance. It is operational memory. Institutions under stress tend to argue about what happened. A signed, dated, versioned record reduces the number of facts that can be reinvented during conflict.

A practical test for the September process

If the 2026 timetable reaches a special meeting, members should apply a simple test to each proposed resolution. First, identify the observed failure. Second, identify the old clause. Third, state what the new words change. Fourth, identify who gains authority. Fifth, identify the trigger and limit. Sixth, identify the remedy. Seventh, state how the change protects routine registry continuity.

Members should reject package voting where unrelated changes are bundled so tightly that they cannot approve a safeguard without approving a power expansion. Each significant redistribution should be visible. Explanatory papers should distinguish mandatory legal updates, drafting corrections and substantive governance choices.

The voting report should publish more than percentages. It should show eligible voters, participating voters, proxies, abstentions, invalid ballots and the result for every resolution. If exclusions occurred, aggregate reasons and review outcomes should be published without exposing private member data.

After adoption, AFRINIC should publish the executed constitution, filing evidence, effective date and a consolidated redline. It should update operational procedures that depend on changed clauses. Training for election officers, staff and directors should happen before the next contested use, not during it.

Finally, the company should schedule a post-implementation review. A constitution is tested by events. Within a defined period, AFRINIC should report whether vacancy, nomination, meeting and remedy clauses were used; how long they took; and whether disputes were resolved without interruption to services.

That is how "after" becomes evidence rather than branding. The process does not end when members vote. It ends when the new settlement is operative, reviewable and capable of producing lawful governance under the conditions it was designed to address.

The finding: the crisis is the redline

AFRINIC's constitutional history shows real amendment capacity. The 2012 and 2016 texts provide historical baselines. Fourteen resolutions in September 2020 produced a fifth constitution that addressed important corporate machinery. Mauritius law supplied courts and a receiver when internal organs could not restore normal governance. The 2026 review created a route toward another settlement.

None of those facts supports the claim that a post-crisis constitution was already in force by 11 July 2026. The operative comparison remains between pre-crisis texts and the way those texts behaved during crisis. The "after" document was still being made.

That makes the crisis itself the most useful redline. It shows where recent clauses met real events: board vacancies, nomination, elections, member standing, external intervention and continuity. Reform should follow those failure paths clause by clause, not merely replace dates and add broad language about stability.

The standard is executable limitation. A good constitution tells members and officers not only who has power in ordinary times, but what happens when an organ cannot act, how authority returns to normal, how a disputed decision is reviewed and how critical services remain insulated from corporate conflict.

AFRINIC should receive credit when it publishes a clear draft, answers comments, holds a valid vote and files the adopted text. It should receive stronger credit when the new rules work. Until then, consultation is evidence of intent, not completion.

The lesson reaches beyond one registry. Institutional legitimacy is not created by a new version number. It comes from a traceable chain: observed failure, precise amendment, authorised adoption, legal effect, constrained exercise and practical remedy. AFRINIC's next constitution can meet that standard. The public record as at 11 July 2026 shows that the work had begun, and that it was not yet finished.

The later evidence will be unusually easy to test. Either a final text was published or it was not. Either members received valid notice and an intelligible redline or they did not. Either the required majority approved each substantive change, with a disclosed electorate and count, or the claim of adoption will remain incomplete. Either the executed constitution entered the company's legal record or the website will be describing an aspiration. Those are not hostile questions. They are the ordinary checks that convert institutional narrative into corporate fact.

The deeper measure will arrive only when the rules are used. If the next vacancy, nomination dispute or election challenge can be contained by clear procedures while registry services continue, the reform will have answered the crisis. If the same questions return to court because no organ can conclusively identify the lawful next step, another edition will have changed the document without changing the institution. AFRINIC's members should therefore judge the new constitution twice: once at adoption, and again at its first serious test.