AFRINIC’s Mauritius Choice: Neutral Venue or Legal Single Point of Failure?

Summary

• AFRINIC's founding application mapped four different control locations: legal and administrative functions in Mauritius, registration and engineering in South Africa, backup in Egypt and training in Ghana.
• That distribution reduced technical and staff-concentration risk, but it did not distribute the company itself: registered office, board validity, membership powers, assets and court remedies remained tied to one host jurisdiction.
• The later receivership sequence demonstrates the consequence of that legal concentration, not the founders' motive and not proof that the risk was predictable in 2004.
• The defensible finding is double-sided: a single legal home made AFRINIC enforceable and preservable, while also making Mauritian corporate law the route through which continental registry continuity had to be protected when governance authority failed.

Four places, two kinds of risk

The most useful way to begin AFRINIC's Mauritius question is not with a court order two decades later. It is with a four-location control map from the founding recognition materials. The Updated AFRINIC Application for Recognition described a design in which legal and administrative functions sat in Mauritius, registration and engineering work sat in South Africa, backup capacity sat in Egypt and training activity sat in Ghana. The map matters because it shows that the founders did not imagine resilience as a single office, a single city or a single technical facility.

That map divided operational failure from legal failure. A technical outage in one site could be softened if backup systems were available elsewhere. A staffing or skills gap could be reduced if training had an institutional base beyond the legal headquarters. A registration function could be placed near a pool of operational skill without making that place the corporate seat. In that sense, the plan was not geographically naive. It recognised that an African regional registry would need more than an address in an incorporation file.

The same map also shows the boundary of that distribution. Corporate existence was not divided among four countries. The legal person, the place where formal papers could be served, the company structure and the forum for statutory remedies were concentrated in Mauritius. Technical continuity and corporate continuity were therefore not the same thing. AFRINIC could spread servers, engineering habits and training across different places while still depending on one legal jurisdiction for the company that held the regional registry role.

That distinction is the core of the article. A regional Internet registry administers a unique resource ledger for a service region. The ledger needs continuity even if the operator faces governance stress. Yet the operator also needs legal personality, contracts, staff arrangements, assets, members, directors and enforceable duties. The founding design reduced some risks by distributing work. It accepted a different kind of concentration by placing the company inside one host law.

The question is not whether Mauritius was a safe or unsafe choice in the abstract. The available record does not contain the host-country bids, scoring matrix, founders' legal advice, tax analysis, liability advice, original incorporation certificate or full original constitution. Without those records, motive cannot be reconstructed responsibly. The question that can be answered is narrower: what did the design distribute, what did it centralise, and how did later Mauritian proceedings reveal the practical meaning of that centralisation?

Seen that way, the title's alternative is intentionally unsettled. Mauritius may have offered administrative credibility and a workable legal form. It may also have become a single legal control point when board authority and corporate continuity were disputed. Those statements are compatible. A legal home can be useful because it creates a forum and a remedy. The same forum can become decisive for an entire region when the company is the one recognised registry.

Reconstructing the choice from incorporation, not crisis

The chronology has to start with incorporation. The final application dates AFRINIC's incorporation in Mauritius to February 2004. That date precedes provisional recognition and final recognition. It means the corporate vehicle existed before the global coordination system fully accepted it as the African regional registry. The company came first; the completed recognition sequence came later.

That order is important because it prevents a retrospective reading in which the later receivership defines the original decision. In February 2004, the available record shows incorporation in Mauritius and a wider operational design across several countries. It does not show that the founders anticipated the exact governance failures that later required court involvement. It also does not show that they were indifferent to concentration risk. The distributed operating design is evidence against a simple story that all authority and capability were placed in one site.

Provisional recognition was a separate event. The 30 September 2004 provisional recognition date belongs to the recognition process, not to the act of forming the company. ICANN's 11 October 2004 notification letter then identified AfriNIC Ltd as a Mauritius company limited by guarantee under the Companies Act 2001. That letter matters because it confirms how the emerging registry was presented to the wider Internet coordination system: not merely as a regional committee, but as a company with a specific host law.

The distinction among February incorporation, 30 September provisional recognition and the 11 October notification is not a technicality. Incorporation gave the project a legal body that could hold responsibilities and engage with other institutions. Provisional recognition indicated that the recognition authority was prepared to move toward acceptance under conditions. The notification recorded the form of the entity and the legal framework in which it existed. Each step carried different consequences.

Final recognition then came after further assessment and transition work. It would be misleading to collapse these events into one founding moment. By the time final recognition arrived, the company was already a Mauritius legal entity and the operating plan had already assigned functions elsewhere. That sequence matters for accountability because corporate form shaped who could sign, be sued, hold assets, hire, contract and be subject to host-law duties while the registry role was still being secured.

The available record therefore supports a precise claim and blocks a broader one. The precise claim is that Mauritius was the legal home selected for the company while other work was deliberately placed in South Africa, Egypt and Ghana. The broader claim that cannot be made is that Mauritius was selected because the founders sought a neutral venue, a weak venue, a protective venue or any other specific legal advantage. The record supplied for this analysis does not disclose that reasoning.

This limit is not a minor caveat. Host-country selection can involve many considerations: company law, political stability, taxation, ability to operate, regional acceptability, travel, bank access, language, communications, cost, public administration and legal familiarity. Some may have favoured Mauritius. Others may have favoured alternatives. Without the comparative record, the correct method is to describe the legal consequence of the selection, not to invent the motive.

The strongest defence of one legal home

A single legal home is not an institutional defect by itself. Every incorporated body needs a law under which it exists. A company that exists nowhere has no stable personality, no clear registered office, no settled duties for directors, no court forum for members and creditors, and no predictable path for asset preservation. The choice of one host can therefore be a discipline rather than a weakness.

For a regional registry, that discipline has real value. The registry needs to sign agreements, employ staff, receive fees, keep accounts, preserve records and hold assets. It must be able to act continuously even when individual officeholders change. A legal form gives outsiders and members a way to understand who the institution is. It also gives courts a way to preserve assets if internal organs fail.

The company limited by guarantee model fits part of that logic. The bundle identifies AfriNIC Ltd as a Mauritius company limited by guarantee under the Companies Act 2001. Such a form can suit a non-shareholder membership body because the institution is not presented as an ordinary equity vehicle. The registry role is not to distribute profits from number administration; it is to maintain a public-interest function under member and governance rules. The selected materials do not provide the original guarantee terms, so this article cannot describe the member guarantee in detail.

The point is narrower: the form supplied a corporate body for a registry function.

One forum can also improve enforceability. If members dispute whether directors are validly in office, if assets need protection, or if a receiver must be appointed, a known host law gives the dispute a place to be heard. Fragmented legal identity could create a different risk: parties might argue over which court could act, which statute governs duties, and whether an order in one country controls assets or records in another. A single legal home can reduce that uncertainty.

This is the strongest host-law defence and it deserves full weight. The fact that Mauritian courts later became important does not automatically prove that Mauritius was a bad choice. It may show that the host law supplied a remedy when corporate organs failed. A legal system that can preserve assets and push a company back toward valid governance may be a safeguard, not merely a vulnerability.

The defence is especially strong when contrasted with technical single-site risk. The founding plan did not rely on one physical location for all activity. It separated legal and administrative work from registration and engineering, backup and training. That is a rational resilience pattern: make the company legally coherent, but distribute the operational capabilities that could be disrupted by local failure.

The limitation is that the legal body itself remained indivisible. If the legal person is central to contracts, assets, member authority and recognition, then any deep corporate dispute can become a service-region problem. The remedy may be available, but the need to use the remedy still means one host jurisdiction has become the control channel through which continuity must be defended. That is the tradeoff, not a moral verdict on the host.

From registered office to remedy

The remedy path begins with legal personality. AFRINIC needed to be more than a loose network of technical supporters. Legal personality allowed it to exist as a body distinct from any one founder, employee, director or member. It could hold assets, enter arrangements and persist across changes in personnel. That persistence is essential for a registry because the number-resource record must continue even when governance changes.

The next step is the registered office. The current AFRINIC Bylaws fix the registered office in Ebene, Mauritius and place the company under the Mauritius Companies Act 2001. Current text is not proof of the exact 2004 constitution. It is, however, evidence of the continuing legal architecture: a Mauritius company, a registered office, members, directors and amendment powers organised through one corporate framework.

The Mauritius Companies Act 2001 supplies the statutory environment for companies, directors, members, remedies, receivership and continuity. A statute of that breadth should not be treated as if every section automatically applied in the same way at every later dispute. Versions, pleadings, orders and judicial reasons matter. Still, the existence of a host-company statute is what makes the remedy path intelligible. The company is not floating above law; it is situated inside one legal system.

From the registered office, the path moves to corporate organs. A company acts through directors, members and authorised officers. If board authority is valid, corporate decisions can be made in the ordinary way. If board authority is lost, disputed or unable to meet quorum, the question becomes who can preserve the company and how authority can be restored. That is not a routing question. It is a corporate-law question.

The bundle's later institutional record states that AFRINIC lost board quorum and that a 12 September 2023 receivership judgment followed. This article does not treat that summary as the full reasoned judgment. It uses it for the limited proposition it supports: the internal governance condition became serious enough that Mauritian legal remedies were invoked to preserve or administer corporate continuity. The details of holdings, party arguments and procedural steps require the full court file.

The path then moves from organs to assets. A registry has assets that matter for continuity: funds, contracts, records, systems, equipment, staff obligations and claims. If directors cannot validly act, assets may need a legally recognised custodian or court-supervised officer. Receivership is therefore not merely an outside intrusion. In some circumstances it is the method by which a company can be kept from paralysis while its governance problem is addressed.

The path also passes through members. A registry company tied to membership governance cannot be preserved only by locking assets in place. It must regain a valid route to member participation, election or governance renewal. The bundle's proceeding update describes an election mandate in connection with the restored receivership result. That is a limited procedural description, not a complete account of the legal reasoning. It does show that the host-law remedy was connected to reconstituting corporate authority, not only to holding property.

Finally, the path reaches corporate continuity. The point of intervention is not to replace the unique registry ledger with a new one. It is to preserve the company that has the recognised registry role while valid governance is re-established. That is why legal concentration matters so much. The continuity of the service region depends on a company whose corporate distress must be managed through the host forum.

What the distributed plan could contain

The four-location plan could contain certain kinds of operational risk. If a registration or engineering office faced local disruption, the existence of backup capacity in another country could reduce the danger of data loss or service interruption. If an early staff group lacked enough experience, training functions could build capacity outside the legal headquarters. If one city became impractical for a particular activity, other sites could hold knowledge and infrastructure.

That distribution also helped institutional legitimacy. A continent-wide registry could not easily ask for trust if every visible function sat in a single place. Putting legal administration, technical activity, backup and training in different countries signalled that the project was not simply a local company claiming continental authority. It created a broader operating footprint, even if the company itself remained incorporated in one jurisdiction.

The plan could also reduce perception risk. Regional infrastructure institutions often face suspicion that one national cluster will dominate. A multi-site design can answer part of that concern by making different countries relevant to the institution's functioning. Mauritius held the legal and administrative role. South Africa held registration and engineering. Egypt held backup. Ghana held training. The exact weight of each role cannot be measured from the provided materials, but the allocation itself is clear.

Yet the distributed plan could not contain every failure mode. It could not divide the legal person into four separately valid companies without changing the nature of the institution. It could not make every host court simultaneously responsible for board validity. It could not make a technical backup site decide who had authority to instruct staff, sign contracts or organise elections. Those questions are anchored in corporate law.

The plan therefore created a useful split: spread operational capability, concentrate legal personality. That split makes sense if the main feared risks are technical outage, local disruption, early skills shortage and legitimacy optics. It is less protective if the central risk becomes invalid governance, board paralysis, membership dispute or asset preservation. Those risks follow the company, not the server.

This distinction explains why the later legal events are analytically useful but not morally decisive. A receivership stress test does not mean the operating distribution failed. The distribution was not designed to make Mauritian corporate remedies unnecessary. It was designed to let the registry operate across more than one physical and technical base while the company remained legally coherent. The later events tested the legal coherence side of the design.

That is also why the phrase "single point of failure" has to be used carefully. In engineering, a single point of failure is a component whose failure can disable a system. In corporate governance, the analogy is imperfect. Mauritius did not simply fail and disable the registry. Rather, Mauritian law became the route through which failure inside the company had to be addressed. The risk was not only breakdown; it was dependency on one forum for repair.

The later stress test

The later proceeding sequence should be used as a stress test, not as a complete litigation recap. The supplied records identify three distinct markers. First, the annual institutional record describes a 12 September 2023 receivership judgment after loss of board quorum. Second, the Court Update reports that a 15 October 2024 appellate result restored the 12 September 2023 judgment, the Official Receiver and an election mandate. Third, the fixed chronology identifies the February 2025 appointment of a replacement receiver as a later separate act.

Those markers show that Mauritian legal institutions became central to corporate continuity. They do not establish every holding. They do not settle every pleading. They do not identify every party position. They do not prove the entire procedural status beyond the limited descriptions supplied. Treating institutional summaries as if they were full reasoned judgments would overstate the record.

The 12 September 2023 marker is still important. A receivership judgment is a major corporate continuity event. It indicates that ordinary governance mechanisms were not sufficient to keep the company functioning in the normal way. For a regional registry, that condition has significance beyond internal company administration because the company is tied to number-resource service. The legal remedy did not concern a random company; it concerned the corporate body through which a regional registry role was exercised.

The 15 October 2024 marker is also important because it describes appellate action that restored the earlier judgment and the Official Receiver. Again, the record available here is not the full judgment. The limited point is that appellate proceedings could alter who had recognised authority to preserve assets and move toward elections. That places continental registry continuity inside a host-country appellate process.

The February 2025 marker must be kept separate. The appointment of a replacement receiver is not the same event as the original receivership judgment or the 2024 appellate result. It is a later step in the remedial sequence. The distinction matters because collapsing the dates into one story would make the legal path seem simpler than it was. Corporate continuity can require multiple acts, each with its own status and limits.

Together, the markers demonstrate consequence. They show what it means for a regional registry's corporate body to live under one host law. When board authority and continuity are under strain, the host forum becomes the place where preservation, election mandates and receiver authority are addressed. That is legal concentration in action.

They do not demonstrate foreseeability. The record does not show that a founder in February 2004 predicted a future loss of board quorum, a 2023 receivership judgment, a 2024 appellate result or a replacement receiver in 2025. Later stress can reveal a mechanism without proving that the mechanism was obvious at formation. The difference between consequence and foreseeability is essential to a fair analysis.

They also do not prove that another host would have avoided the crisis. Any incorporated body in any jurisdiction would need rules for directors, members, assets and remedies. A different host might have supplied faster remedies, slower remedies, weaker remedies or stronger remedies. Without a counterfactual comparison of rejected host jurisdictions and their company-law tools, superiority cannot be asserted.

Legal concentration did not cause everything

The fact that legal remedies became central does not mean legal concentration alone caused AFRINIC's governance crisis. A company may have one legal home for decades without board paralysis. A multi-country operating footprint can still function under one corporate law. A receivership event usually reflects internal governance conditions as well as the availability of statutory remedies. The host forum is the channel through which the crisis is managed, not automatically the origin of the crisis.

This distinction matters because it prevents a simplistic host-country morality tale. If Mauritius had supplied no remedy, the critique might be that the registry lacked an effective preservation mechanism. If Mauritius supplied a remedy, the critique cannot be merely that the forum acted. The serious question is whether the concentration of legal authority in one jurisdiction made the entire region dependent on that forum once internal governance failed.

The answer is yes in a limited sense. Board validity, receivership, corporate assets and election-related remedies were not distributed across the same four-location map as registration, engineering, backup and training. They were located in the corporate framework of the Mauritius company. That made the host jurisdiction a control point for recovery.

The answer is no in a broader causal sense. The available record does not prove that Mauritius caused the loss of quorum, caused the disputes, or made a later crisis inevitable. It does not prove that another jurisdiction would have produced better member governance. It does not prove that the host law was unusually risky when chosen. Those claims require records not included here.

The balanced conclusion is that legal concentration magnified the public significance of internal governance failure. Once the recognised registry body had a corporate crisis, the remedy path ran through Mauritius. The host law did not have to be the cause of the problem to become the decisive repair channel. That is often how institutional design works: a choice made for coherence becomes visible only when stress reveals what was centralised.

The same reasoning applies to institutional legitimacy. The multi-site operating design supported regional legitimacy by distributing visible functions. But legal legitimacy rested on the company and its rules. If members believed those rules were not functioning, the remedy still had to be pursued through the host-law structure. Operational legitimacy and corporate legitimacy can diverge.

The policy implication is not that every regional body should avoid a single legal home. That is impossible for most corporate forms. The implication is that a registry whose function is unique should treat host-law concentration as a disclosed continuity risk. Members should know which failures are handled by technical redundancy and which failures require court or statutory remedies.

The current Bylaws and the problem of historical proof

The current Bylaws are useful because they show the living corporate structure: Mauritius law, a registered office in Ebene, membership arrangements, board powers and amendment mechanisms. They are not a time machine. Current consolidated text may differ from the original 2004 constitution. Amendments, practice and court interpretation can change the legal environment over time. A founding analysis must keep that limit visible.

The Companies Act has a similar evidentiary character. It supplies the host statutory framework for companies and remedies, but a general statute is not the same thing as a decided case. To know exactly how a provision applied in a given dispute, one needs the relevant version, pleadings, orders and reasons. This article therefore treats the statute as the legal environment and does not convert it into unsourced holdings.

The provisional recognition letter also has a narrow role. It confirms that ICANN identified AfriNIC Ltd as a Mauritius company limited by guarantee under the Companies Act 2001. It does not supply the incorporation certificate. It does not explain why Mauritius was chosen. It does not describe legal advice. It does not compare alternative jurisdictions. It is a notification, not a founding legal memorandum.

The final application has a different strength. It is the best supplied record for the original operating distribution. It gives the four-location map and dates the incorporation. But as an application, it is also a favourable presentation by the applicant. It should be used for what it says about design, not treated as an independent audit of motive or risk.

The annual institutional record adds a later chronology of lost board quorum and governance consequences. Because it was produced amid ongoing disputes, it should be treated carefully when legal conclusions are at stake. It can support a limited chronology of institutional events. It cannot replace full reasoned judgments for contested legal propositions.

The combined evidence is enough to identify a mechanism: distributed operations, concentrated legal personality, later host-law remedy. It is not enough to identify all causes, motives or counterfactuals. That is not a weakness of the conclusion. It is the condition that keeps the conclusion bounded.

The missing host-choice record

The most important absent file is the host-country selection record. A serious evaluation would require bids or proposals, criteria, minutes, legal opinions, cost analysis, liability advice, tax analysis and evidence of who participated in the decision. Those materials would show whether Mauritius was selected for neutrality, administrative ease, legal predictability, regional acceptability, cost, diplomacy or some combination of factors.

The second absent file is the original incorporation package. The certificate, original constitution, guarantee terms and founding member arrangements would allow a comparison between the initial legal design and the current Bylaws. Without those records, one cannot say which later governance features were present from the beginning and which were introduced by amendment.

The third absent file is the full court record from 2023 to 2025. The summaries identify major events, but they do not provide every holding, every pleading, every order or every procedural status. A reader should not be asked to infer detailed legal propositions from summary notices. A careful account would distinguish trial-level reasoning, appellate reasoning, receiver powers, election instructions and any later changes in procedural posture.

The fourth absent file is a counterfactual map. To claim that another host would have been better, one would need to compare the company law and remedies of the rejected jurisdictions. That comparison would need to ask what would happen under each system if a registry lost board quorum, assets needed preservation, members needed elections and technical service had to continue. No such comparison is in the supplied record.

These missing files are not rhetorical shields. They are the difference between analysis and speculation. It is possible that the missing record would vindicate the Mauritius choice. It is possible that it would reveal ignored risks. It is possible that it would show that all plausible hosts carried similar legal concentration. The present record cannot choose among those outcomes.

What it can show is that the founders solved one problem and left another inherent problem in place. They solved the problem of over-concentrated operating functions by distributing work. They did not and probably could not solve the problem that a company must have a single legal home. The later stress test made that inherent problem visible.

A bounded counterfactual

A useful counterfactual must begin with constraints. AFRINIC needed legal personality somewhere. It needed recognition by the wider Internet coordination system. It needed operational credibility. It needed records, staff, member governance and continuity. A host choice that made any of those impossible would not have been superior simply because it avoided Mauritius.

The first counterfactual is a different single host. If AFRINIC had incorporated in another jurisdiction, the legal concentration would not disappear. It would move. The relevant question would be whether that jurisdiction offered better remedies, clearer membership protections, faster courts, stronger asset preservation or more suitable company forms. The supplied record does not answer that question.

The second counterfactual is a multi-entity structure. AFRINIC might have imagined separate bodies for legal administration, technical operations, backup and training. But multi-entity structures introduce their own risks: divided accountability, disputes over who controls the registry function, contractual complexity and uncertainty over which body has authority in a crisis. The founding application's simpler design may have been attractive precisely because one company could be recognised while operational work was distributed.

The third counterfactual is a treaty-like or intergovernmental form. That could, in theory, reduce dependence on one company statute. It could also make formation slower, more political and less responsive to the operator community. The supplied record does not indicate that such a route was available or preferred. It therefore cannot be used as an easy alternative.

The fourth counterfactual is stronger internal safeguards within the same host. This is the most practical line of inquiry. A Mauritius company could still have more detailed quorum repair rules, clearer election triggers, stronger member notice provisions, independent continuity planning and pre-agreed asset preservation procedures. The record here does not show whether such protections existed in the original constitution or how they changed over time.

That counterfactual is important because it avoids treating host law as destiny. The choice of Mauritius set the legal environment, but the company's own constitution and governance practice also mattered. Host law supplied the outer framework. Internal rules determined much of the ordinary governance path. The later crisis may therefore reflect interaction between host statute, corporate constitution and institutional conduct.

The bounded counterfactual is this: if AFRINIC had selected another host, the single-legal-home problem would still exist unless the institution adopted a fundamentally different legal form. If AFRINIC had kept Mauritius but designed stronger internal continuity provisions, some later stress might have been easier to manage. Both possibilities require evidence beyond the current record.

Why this mattered beyond Mauritius

The legal concentration mattered because AFRINIC was not just another Mauritius company. It was the corporate vehicle for a continental registry function. The public-interest effect of corporate distress was therefore larger than the private effect inside a normal membership body. When board authority, receivership or election mandates became disputed, the consequences reached network operators across the service region.

That does not mean the host court became an Internet policy body. A court applying company and insolvency remedies is not deciding allocation policy in the ordinary sense. It is deciding corporate authority, preservation and process. But for a registry, those corporate decisions can affect who can act, how assets are controlled and whether governance can be reconstituted. The boundary between corporate law and registry continuity becomes thin under stress.

This is why the distinction between the ledger and its operator matters. The continuity of number-resource records is not identical to the continuity or legal immunity of the company that operates the registry. A unique ledger should not be casually disrupted because an operator has internal problems. At the same time, the operator cannot be placed beyond remedy merely because it performs a public-interest function. Power has to be matched with accountability.

Mauritius became the place where that balance was tested. If courts did nothing, the concern would be that the registry company's assets and governance could remain paralysed. If courts intervened, the concern became that one national forum was controlling the recovery of a regional institution. Both concerns are real. The founding design made them meet in the same place.

For members and resource holders, the practical issue is visibility. They need to know which questions can be answered through registry policy, which through corporate bylaws, which through member action and which through host-law proceedings. Confusing those layers makes accountability harder. A member may have a policy disagreement, a corporate-governance objection or a legal-rights claim. Each travels through a different channel.

The Mauritius choice therefore changed the shape of remedy. It did not make every problem legal. It did not make every legal problem a registry-policy problem. It did mean that when ordinary corporate organs failed, the recognised path for preservation and renewal ran through the host jurisdiction. That is the institutional fact exposed by the later events.

What a mature continuity design would disclose

A mature continuity design for a regional registry should disclose the difference between operational redundancy and corporate remedy. It should say where the systems are backed up, where registration expertise sits, where staff can keep working and how records are protected. It should also say which law governs the company, which court can act, how members can force governance renewal and how assets are preserved if the board cannot function.

That disclosure should not be framed as a confession of weakness. It is normal for a company to have a host law. It is normal for a registry to need legal remedies. The point is to prevent members from mistaking distributed technical functions for distributed corporate authority. A backup site in another country does not answer a quorum failure. A training programme does not appoint valid directors. A registration office does not itself preserve company assets.

The design should also identify the first non-court repair mechanisms. Before a receivership becomes necessary, members and directors should know what happens when board seats are vacant, meetings fail, elections are delayed or quorum is lost. Clear internal steps can reduce the need for external intervention. If court involvement becomes unavoidable, the record should show that internal remedies were attempted or unavailable.

Another disclosure should concern asset and record custody. A receiver or court-supervised officer may be necessary to protect assets, but a registry must also preserve technical and registration records. The company should make clear which records are operational, which are corporate, which are financial and which are member-governance records. Different continuity duties attach to each category.

The design should also separate institutional communications from legal proof. Public updates can keep members informed, but they should not be treated as substitutes for judgments or orders. When litigation affects a registry, members need accessible summaries and also a clear distinction between summary, order, judgment and institutional interpretation. That distinction protects trust because it prevents official notices from carrying more legal weight than they can bear.

Finally, a mature design should map host-law risk alongside technical risk. Technical maps often show data centres, backup sites, security controls and disaster recovery. A governance continuity map should show legal person, registered office, applicable statute, member remedies, board-repair mechanisms, election triggers, asset-protection tools and dependencies on court action. Only then can members see the whole institution.

AFRINIC's founding materials supplied the operational map more clearly than the later legal-risk map. That is understandable for a recognition application focused on whether a new registry could operate. But later events show why the legal map matters. A registry can survive a local technical disruption and still face a company-law crisis that affects its ability to govern itself.

The answer to the bounded question

Why was Mauritius selected as AFRINIC's legal home? The supplied record does not answer the motive question. It proves incorporation in Mauritius in February 2004, identifies the company as limited by guarantee under Mauritius law in the October 2004 notification, and shows legal and administrative functions located there in the founding application. It does not provide the selection matrix or advice needed to explain why Mauritius prevailed over alternatives.

Which operational functions were deliberately distributed elsewhere? The founding application answers that more clearly. Registration and engineering were assigned to South Africa, backup to Egypt and training to Ghana, while legal and administrative functions sat in Mauritius. That allocation reduced concentration in technical capability, disaster recovery and capacity building. It also gave the registry a broader African operating footprint.

How did Mauritian company and insolvency law later become a continent-wide governance control point? The mechanism is corporate continuity. AFRINIC's recognised registry role was exercised through a Mauritius company. When board quorum and authority became contested enough to require receivership and appellate action, the host-law remedy path controlled asset preservation, receiver authority and the route back toward elections. The later events demonstrated that mechanism.

The final answer is therefore neither praise nor indictment. Mauritius supplied the legal person that made AFRINIC enforceable, preservable and recognisable as a company. The multi-country operating design shows that the founders did not concentrate every practical dependency in Mauritius. But legal personality is different from technical resilience. Once the company became the container for a continental registry function, the host jurisdiction became the repair channel when corporate organs failed.

That is not proof that Mauritius was a bad choice, a neutral choice or the cause of later crisis. It is proof that legal concentration has to be analysed separately from operational distribution. AFRINIC's founding design distributed work across four places, but it placed corporate remedy in one. Later proceedings made that single legal channel visible. The institutional lesson is that a registry can spread its operational surface and still owe its deepest continuity remedy to the law of one host.