Summary

  • AFRINIC’s official case index identifies Cloud Innovation Ltd vs AFRINIC 103/2021 in the Supreme Court of Seychelles as an application for a perpetual injunction, entered on 3 November 2021, served on 20 December 2021, later marked completed, and recorded as “Application set-aside.”
  • The available official record does not establish the set-aside date, the court’s reasons, the precise operative terms, costs, an appeal route, or a ruling on the underlying contract, membership, number-resource, ownership, leasing, routing, damages, policy, or final-entitlement questions.
  • “Perpetual” describes the relief sought or ordered. It cannot exempt a legal instrument from review, variation, discharge, appeal, or set-aside, and it cannot make one procedural state a permanent source of registry authority.
  • The sound response is exact obedience to any effective court order combined with exact status tracking, reversible registry action, continuity protection, and a public legal-state receipt that distinguishes the application, any order, its actual holdings, and its current status.

The two words after the grand adjective

The official AFRINIC case index gives this story its entire factual spine. In row 30, it identifies Cloud Innovation Ltd vs AFRINIC 103/2021 in the Supreme Court of Seychelles. The entry is dated 3 November 2021. It describes foreign proceedings and an application for a perpetual injunction. It says service was effected on 20 December 2021. It later marks the matter completed and records the outcome as “Application set-aside.” The index also links to a ruling, but the evidence available for this analysis does not include a full transcription of that ruling.

Those details are enough to pose a serious institutional question, but not enough to answer every legal question that readers may bring to the dispute. Cloud Innovation Ltd and AFRINIC were the opposing parties named in a Seychelles proceeding. The relief was described in unusually durable terms. Yet the recorded disposition says that the application was set aside. That combination matters because AFRINIC is not an ordinary commercial counterparty whose internal records affect only itself. It maintains technical registry functions whose accuracy, continuity and neutrality matter to network operators and resource holders.

A legal direction touching such an institution can influence decisions about records and services even when the underlying dispute remains contested.

The word “perpetual” can therefore mislead twice. First, it can make an application sound like a final adjudication when the official entry identifies an application and does not, on its face, tell us exactly what relief was granted. Second, it can make any resulting restraint sound immune to later legal change. The index defeats that second intuition in the bluntest possible way: the matter was completed, and the application was set aside. Permanence in the name of a remedy was not permanence in the life of the proceeding.

Precision begins by resisting the temptation to turn those two words into courtroom theatre. The entry does not say why the application was set aside. It does not give the date of that disposition. It does not disclose the precise operative language, any costs decision, or any route of appeal. It does not establish whether the court first granted a perpetual injunction and later displaced it, or whether the set-aside concerned the application in exactly the literal sense used by the index. It does not establish what AFRINIC did after the disposition or what happened on any network.

A linked ruling exists, but without a verified transcription here, its reasons and language cannot responsibly be supplied from imagination.

That restraint is especially important because nearby proceedings can create an illusion of completeness. Another Seychelles matter, MA 239/2021, is separately listed as connected to CS 103/2021. Its later disposition and its reasons belong to a different legal record. They cannot be carried across the case boundary to fill gaps in this one. Nor can events in Mauritius be used as a substitute for what the Seychelles court said in CS 103/2021. Similar parties, connected disputes and adjacent dates do not merge separate instruments into a single judgment.

The right opening question is consequently modest: what does this record actually establish? It establishes the identities of the parties and the court, the case number, the application type, the entry and service dates, the classification as foreign proceedings, the completion marker, and the literal recorded disposition. It also establishes a temporal lesson. The legal state associated with the word “perpetual” was not safe to treat as an unchanging foundation for registry governance. Everything beyond that must be proved separately.

Procedure is not substance, and neither is later status

The clearest way to read the proceeding is along three tracks that must never be collapsed.

The first is procedure. What was filed? In which court? Against whom? When did the case appear on the index, and when was service effected? What form of relief did the application seek? What later procedural description appears in the official record? For CS 103/2021, the index answers several of those questions. It identifies an application for a perpetual injunction in the Supreme Court of Seychelles, records an entry date of 3 November 2021, records service on 20 December 2021, and ends with “Application set-aside.”

The second track is substance. What underlying legal rights did a competent court actually decide? Did it resolve any dispute over a resource services agreement, membership, registered number resources, ownership, leasing, customer geography, routing, damages, policy validity, registry authority, or final entitlement? The record available here does not establish a holding on any of those questions. An application-level disposition must not be expanded into a merits judgment simply because the wider conflict made those subjects politically or commercially important.

The third track is later status. Even when an order is made, its current force cannot be inferred from the adjective used to describe it at the beginning. Was it stayed, varied, discharged, appealed, replaced, or set aside? Did a superseding instrument change what parties were required to do? Did the registry act in reliance on it, and if so, was that action later corrected? The case index supplies one later-state signal—“Application set-aside”—while leaving the date, reasons, precise consequences and any ensuing registry conduct unestablished.

Each track answers a different question. Procedure tells us which legal vehicle moved and how the docket describes its fate. Substance tells us which contested rights, if any, were adjudicated. Later status tells us what legal instrument remained effective at a given time and what institutions did in response. A reader who jumps from the first track to the second invents a merits victory. A registry that jumps from the first track to the third risks preserving an obsolete operating state. An advocate who selects only one date may describe a legal reality that was accurate for a time but false later.

This three-part discipline does not diminish the court. It is how the court’s authority is respected. A judicial command should be followed according to its actual terms, by the parties it actually binds, for the period in which it is effective. That is a stronger conception of legal obedience than treating a case label as a free-floating permission. It also prevents a private institution from attaching its preferred institutional story to a judicial act that may have said much less.

Consider what the phrase “application for a perpetual injunction” does not answer. It does not say whether relief was granted. If relief was granted, it does not say what conduct was restrained, which parties were bound, where the restraint operated, when it became effective, or on what conditions. It does not identify the questions expressly decided or reserved. It does not tell us whether later review changed the instrument. Those are not technical niceties. They determine what lawful compliance requires and where lawful compliance ends.

The discipline is equally important in the other direction. “Application set-aside” does not establish that every allegation by the applicant was rejected on its merits, that every position taken by the respondent was approved, or that either party acquired final entitlement to the contested resources. A set-aside can be consequential without being comprehensive. The exact consequence depends on the exact ruling. Where that ruling’s full text has not been verified, responsible analysis must stop at the boundary of the official entry.

The strongest case for durable relief

There is a serious benign argument for perpetual injunctions. Recurring interference can produce recurring applications, repeated uncertainty and repeated costs. Relief designed to endure may give parties a stable legal position and reduce the need for emergency litigation. If a court has issued a valid operative order, the parties should obey it. A number registry cannot treat a judicial command as optional merely because an appeal, variation or later application is conceivable. Improvising around the order could create its own legal and operational disorder.

That argument deserves to be stated without caricature. Stability is valuable. Network-related administration depends on reliable records and predictable service. A restraint that lawfully prevents repeated disruption may protect both. The prospect of later review does not suspend every court order automatically, and the possibility of reversal is not a licence for institutional defiance. If the order binds AFRINIC, AFRINIC must know what it requires and comply while it remains operative.

But the benign case leads to classification, not mythology. Because the order matters, its exact state matters. Because compliance matters, the boundary of compliance matters. Because stability matters, the institution must be able to move from one verified legal state to the next without converting a reversible judicial position into irreversible infrastructure change. The answer to legal uncertainty is a better account of legal status, not a larger claim of registry power.

“Perpetual” has a proper legal function. It can distinguish relief intended to continue from relief designed to expire quickly or preserve a position only until another stage. But it is not a charm against judicial process. An enduring form of relief can still be varied, discharged, appealed, superseded, or set aside through law. The relevant question on any later day is therefore not whether the original description sounded permanent. It is which instrument was effective on that day, what it actually required, and whether a later event changed it.

This distinction is particularly important when an institution controls records or services that others may mistake for the underlying rights themselves. A registry may be asked to change a registration, suspend a service, preserve a status, or refrain from action. If it treats the adjective attached to relief as a permanent delegation of power, it may continue an action after the legal basis has changed. If it treats every contested order as meaningless, it may fail to comply when compliance is required. Both errors arise from imprecision. Exact legal-state tracking avoids both.

The institutional lesson of CS 103/2021 is thus not that perpetual injunctions are weak. It is that their force is legal rather than magical. Their durability depends on the current state of the legal instrument. The official index’s final notation is powerful precisely because it changes the classification that a reader might otherwise derive from the original label. The record begins with a claim to enduring restraint and ends with a procedural disposition that cannot be ignored.

A court’s force does not become a registry’s sovereignty

NRS describes AFRINIC as a private, member-based number registry. Heng Lu’s analysis supplies the controlling institutional distinction: a registry is a technical bookkeeper, service provider and coordinator, not a sovereign. Its legitimate functions are narrow but important. It preserves uniqueness, maintains accurate and auditable registration information, supports contactability and security-relevant coordination, records recognised changes, and helps sustain continuity. Those functions make its records useful. They do not transform those records into declarations of public title or the registry into a government.

That distinction clarifies what happens when a court order enters the picture. A court may have lawful power to require an identified party to act or refrain from acting. The command’s force comes from the court and applicable law. AFRINIC may be bound as a party. It may have to implement a carefully defined measure. Yet it does not acquire the court’s authority by complying. It cannot carry judicial force into unrelated decisions, enlarge the order beyond its terms, or treat a moment of court-backed action as proof that it possesses legislative, regulatory, police, prosecutorial, punitive, confiscatory, or public-law adjudicative power.

This is not semantics. The difference determines who must justify coercion, how far an action can reach, and what happens when the legal basis changes. If AFRINIC acts because a court has ordered it to do something, its public account should identify the effective instrument and the measure tied to it. If the instrument is varied or set aside, the dependent operating state should be reviewed without delay. The institution should not say, in effect, that a court once touched the dispute and therefore the registry may now decide the rest.

Nor does a registry entry settle the physical or commercial world by itself. A record may describe who is registered for a resource or which contact details are associated with it. It may be essential to coordination. But description is not sovereignty. A registry database cannot adjudicate ownership of businesses or networks, determine damages, decide the validity of every contract, or make a disputed policy position equivalent to law. Those propositions require their own competent legal basis.

The temptation to inflate authority is strongest when a court order appears to favour the institution’s desired operating state. The registry can then present compliance as vindication of its broader mandate. The opposite temptation arises when an order is inconvenient: the institution may minimise the court or emphasise its own technical discretion. Both temptations should be rejected. The stable rule is symmetrical. Courts supply judicial authority within the scope of their instruments. The registry supplies neutral technical administration within the scope of its legitimate role.

Neither role should be rhetorically transferred to the other.

Heng Lu’s non-punishment principle sharpens the point. A private registry should not use its control of records and services to punish, confiscate, prosecute, or discipline as though it were a public authority. It may carry out a lawful court command. It may enforce genuinely contractual terms through proper processes. It may maintain technical integrity. But it should not turn those bases into an independent power to impose consequences that the court did not order and the contract does not support. When the legal instrument later changes, every measure that depended on it must remain capable of review and correction.

The set-aside entry matters because it exposes the danger of borrowed permanence. Suppose an observer saw only “perpetual injunction” and assumed that the registry’s preferred status had become permanent. That observer might treat later database choices as settled governance rather than contingent compliance. Yet the official index itself records a changed state. The court’s procedural authority must be followed through that change. A registry cannot preserve the aura of the first label after the official record has moved on.

What the official record proves—and what it cannot carry

Official sources deserve weight, but weight is not the same as unlimited scope. AFRINIC’s case index is the strongest available source for what AFRINIC publicly recorded about this proceeding. It identifies the case, the parties, the forum, the application, the chronology of entry and service, the completion state and the set-aside notation. The linked PDF is the selected primary ruling location. In the absence of a verified full transcription here, neither source can responsibly be made to speak the missing reasons.

The index is also an institutional publication by a party to the dispute. That does not make its docket facts unusable; it means their role should be defined. The entry can prove what the index says. It cannot independently prove AFRINIC’s merits position, and it cannot substitute for the court’s full reasoning. The prudent reader uses the official entry as a receipt, not as an invitation to infer every legal consequence that would flatter one party.

NRS, Heng Lu, LARUS and BTW occupy different evidentiary positions. NRS helps explain AFRINIC’s private member-based character and why member rights and the practical ability to leave matter to institutional legitimacy. Heng Lu explains the boundary between bookkeeping and sovereignty, the prohibition on private punishment, and the need to protect continuity while disputes remain live. LARUS describes the operator-facing burden of documentation, delay, contested registry status and legal conflict. BTW provides a model for due process, review, stays and reversible operating state. None of these sources issued the Seychelles disposition.

None supplies the missing date or reasons. Their value lies in explaining why the proven procedural facts matter.

This division of labour is more rigorous than either official deference or official suspicion. Official deference would treat AFRINIC’s index as the entire truth of the dispute. Official suspicion would discard it because AFRINIC was a party. A better method asks what each source is competent to establish. The court record bears the case identity and legal event. Institutional analysis explains the registry’s role. Operator evidence identifies the continuity channel. Due-process analysis shows how systems can respond without irreversible overreach.

Uncertainty must remain visible through the analysis. The exact set-aside date is unknown on the record used here. The grounds are unknown. The precise operative wording is unknown. Costs and any appeal route are unknown. Whether a perpetual injunction had first been granted is not established. No post-disposition registry action or network effect is established. There may be fuller or later public records elsewhere, but their absence from the material examined is not proof that they do not exist. It simply limits what can be said here.

This last point protects readers from a common evidentiary error: treating absence as a finding. “The record does not establish X” is not the same as “X never happened.” It is a statement about the available denominator. The distinction is essential in legal reporting, where a missing ruling, an incomplete index or a neighbouring case can tempt a neat but unsupported narrative. The honest conclusion may be narrower than the desired one, yet it is more useful because future evidence can update it without having to dismantle invented certainty.

Why a cross-border order touches running networks

The case index labels CS 103/2021 as foreign proceedings. That description matters not because it proves a jurisdictional flaw—the record establishes no such reason—but because it highlights a practical translation problem. Courts act through legal instruments addressed to parties. Registries operate technical and administrative services relied upon by people and organisations beyond the courtroom. Moving from the first domain to the second requires exact interpretation.

A number registry’s actions can alter registration records, contactability, access to administrative services and the continuity assumptions on which operators rely. LARUS’s account of the wider conflict identifies documentation burdens, delay, contested registry status and legal uncertainty as real operator-facing costs. That account does not prove a particular loss in CS 103/2021. It explains the mechanism through which imprecise legal-state handling can become operational friction.

The mechanism can be described without inventing an outage. If a legal instrument appears to require a change, staff must identify the resources, parties, services and time period within scope. Operators may have to produce documents or respond to altered status. Counterparties may hesitate while authority is contested. If the instrument is later displaced, every dependent action must be identified and reviewed. The longer the registry waits, or the more irreversible its intervention, the harder the correction becomes.

These costs can exist even when the underlying network keeps passing traffic. Continuity is broader than packet flow at a particular moment. It includes reliable records, responsive contacts, predictable administration, security coordination, and the ability to demonstrate a legitimate status to counterparties. A dispute that clogs those functions can burden operators without producing a dramatic technical failure. That is why the absence of a proven outage does not make legal-state design irrelevant—and why this analysis must not claim an outage that the record does not show.

Cross-border context raises the value of a common receipt. Staff in one jurisdiction, counsel in another and operators across a service region may encounter the same proceeding through different documents and at different times. If each relies on a descriptive shorthand—“perpetual,” “foreign,” “completed,” “set aside”—they may act on different assumptions. A structured record that preserves the actual instrument, scope and current status offers a common basis for lawful action without pretending to decide questions the court left open.

The role of such a receipt is not to second-guess a court. It is to prevent loss in transmission. A court’s command is authoritative through its legal form. A registry’s operational implementation should retain enough detail to show that the action matches that form. When a later judicial event occurs, the implementation record should reveal which measures depend on the altered instrument. This is ordinary institutional competence, not resistance.

The CS 103/2021 entry captures the risk in miniature. One phrase points toward indefinite restraint. Another records a set-aside. Between them lie the unknowns: what precisely was ordered, when the decisive change occurred, why it occurred, and what followed. An institution responsible for continuity cannot safely fill those spaces with a party narrative. It needs documentary states, dates, dependencies and reversible controls.