Summary

  • On 6 January 2022, the Supreme Court of Mauritius, Commercial Division, set aside the present application in SC/COM/MOT/000382/2021 with costs after reading a letter from Cloud Innovation Ltd’s attorney and taking cognizance of AFRINIC’s stand in another same-day letter.
  • The order gives no policy reasons. It does not identify the relief sought or a policy proposal, reproduce either letter, state a costs amount, or decide the validity, legality, wisdom, representativeness or implementation of a transfer policy.
  • The commissioning context describes a withdrawn and discharged injunction challenge concerning AFRINIC’s transfer policy, but those descriptions are not the court’s words. The order’s operative language is “set aside”.
  • The resulting institutional state is easy to misread: an existing community-policy decision may remain in place after a challenge ends, but procedural survival is not judicial approval and does not turn a private registry rule into law.
  • The strongest benign explanation is not a specific motive but the legitimacy of party autonomy and procedural economy. Because the two letters are unavailable, no choice among settlement, changed circumstances, cost, duplication, evidence or strategy can responsibly be made.
  • The durable remedy is a public non-decision receipt that separates the court ledger from the registry ledger, records what was and was not decided, and leaves competent review available without converting silence into authority.

Read the order before reading a victory into it

The page is almost austere. At its top are the Supreme Court of Mauritius and its Commercial Division. The case number is SC/COM/MOT/000382/2021. Cloud Innovation Ltd appears as applicant; African Network Information Centre (AfriNIC) Ltd appears as respondent. The judge records having read a letter dated 6 January 2022 from the attorney appearing for the applicant. The judge also takes cognizance of the respondent’s stand, conveyed by a letter of the same date. Then comes the operative sentence: “I set aside the present application with costs.” The order was made in Chambers on 6 January 2022 in the name of P. D. R. Goordyal-Chittoo J.

That is a complete account of what the one-page instrument affirmatively supplies. It is also the place where disciplined reading must stop before context is added. The order does not reproduce either letter. It does not state why the applicant’s attorney wrote, what position the respondent took, what relief the application requested, when that application was filed, or what factual and legal propositions had been placed before the court. It names no transfer policy and contains no discussion of policy language. It specifies no amount of costs and says nothing about assessment, allocation beyond the words used, or payment.

Above all, it gives no reasons on policy and no conclusion on policy validity.

The commissioning record connects this application to an injunction challenge concerning AFRINIC’s number-resource transfer policy and characterises the event as withdrawal and discharge. That connection gives the procedural order its subject here, but it cannot rewrite the order. Neither “withdrawn”, “discharged”, “injunction” nor “transfer policy” appears on the page. The judicial verb is “set aside”. The distinction is more than editorial caution. A description carried in contextual records answers what matter observers associate with the docket; the court instrument answers what the judge actually ordered.

Combining those functions carelessly creates a judgment that does not exist.

The narrow policy background can be stated without turning this analysis into a biography of the proposal. AFRINIC’s archive identifies AFPUB-2019-V4-003-DRAFT04 as Draft 4 of the Resource Transfer Policy and records its submission on 5 October 2020. AFRINIC’s Appeal Committee register shows that a 2021 appeal challenged the consensus determination concerning that draft. The appeal filing proves that objections were made; it does not prove that every allegation in those objections was right, and it does not show that a court accepted them. The proposal was later marked expired and archived.

That later status is a separate institutional event, not a retrospective description of what the court did on 6 January 2022.

This limited sequence is enough to expose the central problem. An internal policy decision existed. It was contested through an internal route. A court application associated by the commissioning record with the transfer-policy controversy later ended through a one-page set-aside order. No policy-merits judgment appears in that order. Therefore, the application’s end did not judicially affirm the policy, and it did not judicially invalidate it. The policy-process state could remain intact for the moment simply because the route that might have tested it no longer remained live in that application.

The difference is between continuity and validation. Continuity describes which institutional arrangement is still operative after an event. Validation describes an authoritative decision that has tested and approved a proposition. The first may occur without the second. If an application capable of disturbing an existing state disappears before reasons are given, the state can persist through inertia. Its persistence does not acquire the content of the missing judgment. Silence cannot perform the analytical labour of reasons.

The reverse mistake is just as serious. The set-aside does not prove that the challenge was well founded, that the transfer policy was defective, that the applicant was right, or that the respondent had acted beyond its proper role. Nor does it prove that the challenge was baseless, that AFRINIC’s consensus determination was vindicated, or that the applicant conceded the merits. The court’s page supplies no basis for selecting either story. A non-decision is not a secretly coded decision for the other side.

Costs do not change that boundary. “With costs” is part of the exact disposition and should never be omitted from a faithful account. Yet the presence of costs does not transform a reasonless procedural ending into an adjudication of transfer-policy validity. It does not establish wrongdoing, constitute sovereign punishment, or prove that a particular sum was assessed or paid. Without a costs amount, a later assessment record, or evidence of payment, those entries remain unknown. The responsible reading preserves the words while refusing to add a meaning the words cannot bear.

There is a strong benign case for allowing matters to end without demanding a public merits judgment. Parties control their litigation. They may conclude that an interlocutory contest is no longer the best use of time or resources. Circumstances can change; another procedure or an undertaking may affect immediate needs; proceedings can become duplicative; evidentiary positions can move; costs can rise; legal strategy can change. A private policy institution also needs a workable current state rather than permanent paralysis every time a dispute is initiated. Procedural economy and finality are legitimate values.

But recognising those possibilities is not permission to choose one. The two same-day letters are not reproduced, and the order does not explain their purpose. Settlement, concession, pressure, fear, mootness, error, abuse, changed circumstances and tactical withdrawal are all unavailable as factual conclusions. Even a plausible account would still be invented. The benign case therefore strengthens the argument for restraint: respect the parties’ ability to terminate litigation, preserve the finality of the court’s act, and publish an exact record of the questions that remained unanswered.

That restraint matters particularly because AFRINIC is not a public sovereign. NRS describes it as a member-based number registry. Heng Lu’s analysis places the institutional boundary more sharply: participation can provide technical evidence and practical judgment, but it cannot manufacture authorization from people who did not appoint the participants as their sovereign representatives. A private community process may organise how a registry service is administered. It does not legislate for a continent.

AFRINIC’s legitimate field is consequential but bounded. It keeps number-resource records, helps protect uniqueness, maintains contactability and supports security-relevant coordination. It can record recognised changes of control, administer contractual processes, defend its policies and comply with competent court orders. It is not a sovereign, legislature, regulator, police force, prosecutor, punisher, confiscator or public-law court. The disappearance of an application cannot promote it into any of those roles. A judicial non-decision neither grants new authority nor converts a community rule into public law.

The same separation protects the court. The court need not run a registry in order to decide a legal question properly brought and maintained before it. Its function is not to make allocation records or operate transfer procedures. The registry, in turn, cannot become the interpreter of a missing judgment. The clean institutional position is that the court ledger records a set-aside with costs and no policy reasons, while the registry ledger records whatever policy status, implementation position and later archival change its own evidence supports. Neither ledger should borrow an entry from the other.

For operators, the distinction is not abstract. LARUS explains a general channel by which uncertainty in registry rules and decisions can move into diligence, transaction timing, pricing and continuity planning. A resource holder considering a transfer needs to know what rules a registry will apply. A counterparty needs to know whether the recorded state is stable enough for the contemplated transaction. An operator responsible for continuity needs to distinguish an administrative practice from a court-tested entitlement. If “the application ended” is shortened to “the policy was upheld”, those assessments begin from a false premise.

Nothing in the sealed record proves that this order caused a transfer delay, an outage, a price movement, customer harm or any quantified loss. Those claims would require evidence not present here. The operational importance lies instead in the error channel: ambiguity at the boundary between legal status and private administration can distort decisions even when no particular injury has yet been demonstrated. Precision is preventive infrastructure.

The immediate public task is modest. Quote the operative verb. Attribute the transfer-policy and injunction description to context rather than to the order. Mark policy validity unadjudicated. Keep the later expiry separate. Record the unknowns as unknowns. And place a warning next to any surviving policy state: survival after non-decision is not judicial validation. Those steps do not reopen the application or disrespect finality. They keep finality from being asked to carry a meaning it never acquired.