Summary
- AFPUB-2019-V4-003 began in October 2019 as a Resource Transfer Policy amending section 5.7 of AFRINIC's Consolidated Policy Manual. It addressed inter-regional movement of IPv4 registrations in a system where reciprocity and the receiving registry's rules matter.
- By October 2020 the proposal had reached draft 4. Public records show material questions across versions, including needs assessment, treatment of legacy resources, the roles of source and receiving registries and the procedure for outgoing transfers.
- AFRINIC reported feedback that ARIN and APNIC considered wording in draft 3 incompatible with their inter-regional transfer policies. That feedback demonstrated that local consensus could not make an inter-regional mechanism operational if counterpart rules did not align.
- A 2021 appeal challenged the consensus determination on draft 4 and disputed the treatment and disclosure of material revisions. The appeal's existence proves contested procedure and substance, not that any entity acted in bad faith.
- The proposal was ultimately marked expired and archived. Archival closure removed it from active consideration but did not supply a merits decision on every unresolved issue or eliminate demand for a workable transfer rule.
- A bounded policy process needs a version-by-version disposition, external compatibility checkpoint, fixed decision windows, a law of fatigue and a terminal status that explains whether the proposal failed on substance, support, process or time.
The proposal addressed a market that procedure could not ignore
IPv4 scarcity changed the meaning of registration policy. When unallocated space was plentiful, a registry could concentrate on needs-based issuance and accurate records. As the free pool contracted, organisations sought ways to acquire addresses from existing holders. A transfer policy had to decide whether and how the registry would recognise that movement.
The question was not simply whether private parties could make an agreement. Internet addresses must remain uniquely registered, routing contacts must be accurate, and an inter-regional transaction must satisfy the rules of more than one registry. A transfer that one side recognises and the other rejects cannot produce a coherent global record. Legacy status, needs tests and eligibility conditions can change across the boundary.
AFRINIC already had a policy baseline involving transfers within its region. AFPUB-2019-V4-003 proposed to amend the relevant section and address a wider movement. This placed the African registry inside a global transfer architecture shaped by other regional policies. The community could choose its principles, but it could not make counterpart systems compatible through declaration alone.
The economic stakes increased the cost of ambiguity. Registered IPv4 blocks carried market value and operational utility. Delay could constrain networks seeking addresses or encourage transactions whose registration consequences were uncertain. A permissive rule could also enable resources to leave the region or alter the obligations attached to historical holdings. Both action and inaction distributed value.
These pressures explain why the proposal persisted. They do not justify indefinite deliberation. A serious policy process must convert complex scarcity interests into a decision with known authority, even if that decision is to reject a draft and invite a better one.
Draft 1 opened several constitutional questions at once
AFRINIC's public page identifies draft 1 as submitted on 30 October 2019 and aimed at section 5.7 of the Consolidated Policy Manual. The proposal's title, Resource Transfer Policy, sounded singular, but it implicated several distinct constitutional choices.
Who qualified as the current rights holder? Would AFRINIC test the source, the recipient or both? Which registry's needs policy governed? Could resources move out of the African service region without an AFRINIC needs review? Would legacy resources retain a different status after transfer? What public information would document the transaction? How would a transfer involving a national registry or another RIR be coordinated?
Bundling these choices can make drafting efficient because they interact. It can also make consensus opaque. A entity may support inter-regional transfers while opposing no-needs treatment, or support incoming transfers while fearing unrestricted outflow. One call on the whole draft can conceal the reason people raise or lower a hand.
Draft 1 therefore required an issue map before a consensus map. Chairs needed to identify which clauses were principles, which were operational mechanisms and which depended on external compatibility. Authors needed a revision record showing how each concern changed the text.
The initial version should be judged as an opening proposition, not as a failed final instrument. A difficult policy often requires several drafts. The governance question is whether each draft reduces uncertainty in a visible way or merely moves disagreement into different words.
Inter-regional policy has two sources of authority
An AFRINIC community decision can authorise its registry to act. It cannot bind ARIN, APNIC, LACNIC or the RIPE NCC. Inter-regional transfer therefore requires two compatible sources of authority: the policy of the source side and the policy of the receiving side. Each registry retains its own rules and institutional accountability.
This structure makes reciprocity more than diplomatic courtesy. Suppose AFRINIC says a source need only comply with the receiving registry, while the counterpart requires a particular status or needs test. The texts must align closely enough for both organisations to approve the same event. A clause that appears balanced in isolation may be unusable when applied across rule sets.
Compatibility review should occur early and again after material revision. Staff can send the exact draft to counterpart registries and publish a technical matrix: permitted direction, eligible resource type, source obligations, recipient obligations, legacy treatment, minimum size, holding period, documentation and public record. A response should be tied to a version.
External feedback does not give another registry a veto over AFRINIC's policy objectives. AFRINIC may adopt a rule that operates only with compatible regions or choose terms that counterparts do not yet accept. The community should know the consequence. “Inter-regional transfer policy” is misleading if no counterpart can use it.
This dual authority also complicates consensus. Entities can agree that AFRINIC should permit transfers while disagreeing about the concessions needed for reciprocity. Chairs should state whether the call concerns principle, exact compatible text or implementation with a named set of regions.
Counterpart feedback exposed a real implementation fault line
An AFRINIC Resource Policy Discussion list message in October 2020 reported partial feedback from other registries on draft 3. It said ARIN and APNIC considered wording requiring a source to be the current rights holder and comply with the receiving registry's policies incompatible in particular applications. The exchange also discussed whether another proposal was fully reciprocal and bidirectional.
This feedback was consequential because it tested policy against the organisations that would have to execute it. It moved discussion from abstract fairness to a concrete question: could the same transfer satisfy both registries' conditions? A clause capable of opposite interpretations could delay transactions or create inconsistent records.
The institution needed to do more than quote counterpart concern. It should publish the exact question sent, version examined, full response where permitted, staff analysis and proposed textual remedy. Entities could then distinguish a drafting ambiguity from a fundamental policy conflict.
Timing matters. If external compatibility concerns emerge shortly before a meeting or consensus call, the community lacks adequate opportunity to absorb them. The call should be delayed or narrowed. An operational defect discovered after a consensus declaration can justify return, but the record should explain why it was not tested earlier.
The feedback did not prove that draft 3 or draft 4 was illegitimate. It proved that inter-regional policy required a wider evidence base than internal preference. A proposal may have strong local support and still fail its stated objective because another authorised party cannot act under it.
Draft 4 arrived with a burden of explanation
AFRINIC's archive identifies draft 4 as submitted on 5 October 2020. By a fourth version, entities should not have to rediscover the entire debate. The new text should arrive with a burden of explanation: what changed, why it changed, which objections were resolved, which remained and how compatibility feedback was incorporated.
This burden grows with each version. Early flexibility is productive; later ambiguity is expensive. Entities who reviewed three prior drafts need an efficient way to decide whether their concern still applies. New entities need enough history to avoid repeating arguments without being forced to read every message.
A proper revision table should classify material changes. Needs assessment is substantive. Legacy-resource status is substantive. Moving approval responsibility between source and receiving registries is substantive. Changing a transfer template may be operational but still consequential. A missing item in revision history can undermine trust because entities cannot know whether the consensus call covered the actual policy.
The 2021 appeal later alleged that important changes were not adequately listed. An appeal allegation is not a final finding of fact, but it demonstrates why the burden mattered. When disputed words control valuable resources across registries, version transparency is part of consent.
Draft 4 should therefore be treated as a new decision entity carrying historical context, not as draft 1 with accumulated legitimacy. Earlier support follows only where the policy meaning remains stable and entities can verify that continuity.
Needs assessment was not a minor drafting detail
The appeal materials compare provisions across versions concerning needs evaluation. One form contemplated transfers based on mutual agreement without a general AFRINIC approval, subject to conditions where another region imposed need. Another described an incoming recipient justifying need before AFRINIC, while an outgoing transfer followed the receiving registry's policy.
These are different allocations of authority. A needs test can limit speculative acquisition and preserve an allocation ethic based on demonstrated use. It can also add delay, judgment and documentation cost. Applying it to incoming resources but not to outgoing ones reflects the receiving registry's responsibility for the recipient. Applying counterpart rules to the source can create incompatible or excessive obligations.
Entities could reasonably agree on transfer in principle and disagree sharply on this architecture. Chairs needed to call the issue explicitly. A general statement of consensus on Resource Transfer Policy would not show which needs model had support.
Staff analysis should have described concrete journeys: an AFRINIC member sending to ARIN, an APNIC account holder sending into AFRINIC, a legacy holder, and a recipient seeking a block under each region's rules. For each, identify who evaluates need, which evidence applies and what happens if registries disagree.
This kind of case analysis turns abstract text into inspectable consequences. It also prevents a late claim that a changed clause was merely clarification. Where authority and eligibility move, the policy has materially changed and needs renewed review.
Legacy status carried distributional consequences
Legacy IPv4 resources predate modern regional allocation contracts and can receive different treatment across registries. The question whether transferred legacy resources retain that status affects obligations, fees, documentation and future transferability. It can also redistribute benefits between historical holders and current members supporting registry services.
The public discussion around AFPUB-2019-V4-003 included proposals for transferred legacy resources to lose legacy treatment in certain circumstances. Such a rule can be defended as equalising obligations once a holder voluntarily enters a modern transfer system. It can be criticised as altering a status that one side considers part of the transaction.
Inter-regional movement intensifies the issue because status may not translate cleanly. A resource considered legacy at the source may enter a registry whose policy attaches different conditions. The transfer text must say which institution determines status and when the change occurs. Counterpart compatibility again becomes essential.
The community needed evidence about actual legacy holdings and likely cases, not only principle. How many resources could be affected? Which services are supplied without contract or fee? Would loss of status discourage accurate registration? Would retention create unequal treatment after a market transaction? These questions concern registry sustainability and fairness.
Repeated revision is justified when it resolves such a conflict. It becomes stasis when versions alternate treatments without a clear decision standard. A disposition record should state the competing principles, evidence and chair conclusion rather than allowing legacy status to reappear as a fresh objection at every meeting.
Revision can clarify or move the contested power
Policy drafting often responds to an objection by relocating responsibility. If AFRINIC should not approve an outgoing transfer, perhaps the receiving registry should. If source compliance is problematic, perhaps recipient need should carry more weight. Such changes may solve compatibility. They may also shift discretion without addressing the underlying fairness concern.
Every revision should therefore be tested against a power map. Who decides eligibility? Who can delay? Who holds evidence? Who may reject? Which party can appeal? Which registry updates the authoritative record? A sentence-level comparison may miss that the same practical power moved from one institution to another.
For AFPUB-2019-V4-003, the appeal materials' attention to source and receiving roles shows that entities understood this consequence. The revision history needed to explain not only wording but institutional effect.
This is where chairs should separate drafting from judgment. Authors can propose language. Staff can assess operation. Counterpart registries can report compatibility. The Policy Development Working Group must decide whether the allocation of power is acceptable. Consensus should attach to that decision, not merely to a clean document.
A version that clarifies who acts can be progress even if it narrows the original ambition. A version that uses ambiguous passive language to hide responsibility is not. The history should let readers see which occurred.
Meetings provide intensity, not full continuity
AFRINIC policy discussion occurs on the open Resource Policy Discussion list and at public policy meetings. Meetings concentrate attention. Authors can present a revision, chairs can test understanding, and entities can respond in real time. For a difficult transfer proposal, this intensity can expose disagreements that asynchronous exchange leaves diffuse.
The meeting also creates discontinuity. Attendance changes from one event to the next. Entities may lack time to read a late draft. Remote access and time zones influence who speaks. A show of hands captures the room, not every list entity or affected member. Informal corridor discussion may shape compromise without entering the public record.
When a proposal persists across several meetings, the group at the latest call may differ substantially from the group that developed earlier clauses. Support cannot simply accumulate across rooms as if the constituency were stable. Chairs need a version-specific record integrating list and meeting evidence.
Meeting minutes should state the exact text, principal arguments, questions asked, entity signals and chair conclusion. The list should receive a prompt report and a defined comment period. Material changes made near or during the meeting should not receive instant finality.
A proposal outlives its forum when its formal identity continues while the entities, context and disputed text change around it. The remedy is not to forbid long development. It is to renew authority explicitly at each material transition.
Consensus and exhaustion can look alike
Repeated drafts can reduce opposition because concerns are answered. They can also reduce opposition because entities stop responding. From the outside, both produce a quieter list and fewer objections at the next meeting. Chairs must distinguish convergence from attrition.
Evidence of convergence includes objectors confirming that revisions solved a problem, independent organisations newly supporting the exact text, compatibility tests passing and clear disposition of prior concerns. Evidence of attrition includes falling participation without resolved objections, the same issues resurfacing after long gaps, reliance on absence rather than affirmative reasoning and entities saying that repetition is futile.
No single measure is decisive. A veteran objector may leave for unrelated reasons. A short quiet period after extensive discussion may reflect justified closure. The institution should publish enough longitudinal evidence for its conclusion to be credible.
The longer a proposal remains active, the greater the chair's duty to refresh the record. Summarise prior objections, ask whether they remain, invite new organisations and state what evidence would change the outcome. Do not require every entity to restate the entire case.
AFPUB-2019-V4-003's progression into formal appeal indicates that quiet or meeting support had not produced a universally accepted procedural conclusion. The appeal did not prove consensus absent, but it showed that the distinction between resolution and exhaustion required reasoned review.
The consensus declaration needed a clause-level record
A chair declaring consensus on a fourth draft should identify the clauses or principles that passed, the substantial objections, their dispositions, the participation evidence and the next review stage. A generic conclusion is especially weak where the proposal contains severable issues such as need, legacy status and inter-regional reciprocity.
Clause-level reporting does not turn rough consensus into a vote. It makes qualitative judgment inspectable. The chair can say that broad support existed for inter-regional transfers, that a needs objection was addressed by specified receiving-registry language, that a legacy-status concern remained but did not outweigh the policy goal, and that compatibility depended on counterpart confirmation.
Entities can then challenge the actual reasoning. Perhaps the revised language did not solve the objection. Perhaps an external response contradicted the compatibility finding. Perhaps the remaining concern was repeated but unsupported. An appeal body receives a bounded question rather than a dispute over atmosphere.
Without this record, later readers may treat the entire draft as endorsed or treat one unresolved clause as defeating everything. Both are inaccurate. Consensus can be partial, conditional or sufficient for another stage without being sufficient for implementation.
The proposal's long history made such precision more necessary, not less. Every prior meeting and version created potential confusion about which concern belonged to which text. Clause-level disposition would have converted accumulated discussion into institutional memory.
Appeal is evidence that finality failed, not proof of the winner
AFRINIC's Appeal Committee page lists a 2021 appeal against the consensus determination on AFPUB-2019-V4-003-DRAFT04. The appeal document challenges both substantive changes and the handling of the determination. Its existence is important evidence: at least one formal entity believed ordinary discussion had not produced a procedurally acceptable closure.
An appeal should not be reported as proof that the proposal or chairs were wrong. Appeal systems exist precisely because reasonable people can contest process and judgment. The reviewing body must decide under published standards. The allegation itself identifies where the record lacked shared confidence.
Appeal can improve legitimacy if it freezes the disputed version, states the questions, hears relevant parties, publishes reasons and defines the effect. Does the proposal return to discussion? Does a clause require renewed call? Is the original finding upheld? Are chairs directed to improve reasons? The remedy should match the error.
Delay during appeal has consequences. The proposal remains uncertain, entities may hesitate to engage a parallel draft, and implementation cannot safely proceed. The appeal timetable should therefore be bounded while preserving adequate review. Extensions need public reasons.
Appeal is not an enemy of consensus. It is a check on the authority to declare it. But an institution should study why a proposal reached appeal after several drafts. If material changes or compatibility issues could have been resolved through better records earlier, prevention is preferable to repeated formal review.
The appeal highlighted version-history accountability
Among the appeal's concerns was the treatment of revisions. It compared language across versions and argued that certain changes, including aspects of transfer procedure, were not properly reflected in the revision history. This kind of dispute goes to the entity of consent.
If entities are told that a new draft changes one issue while another material clause also moves, their silence or support cannot safely be attached to the hidden change. A clean text is not enough. Public policy depends on informed comparison, especially for volunteers who cannot reread every section after each revision.
Version accountability has four elements. Preserve every text. Produce a machine-readable and human-readable difference. Classify each material change. Link it to the comment, staff finding, external response or author decision that caused it. The author may make a change not requested by anyone, but should say so.
Chairs should verify completeness before a consensus call. Entities should have a short opportunity to challenge classification. A disputed “editorial” change that affects authority should trigger renewed review. Final status should identify the exact hash or publication record of the text considered.
These controls are not technical ornament. They protect authors from claims that wording changed secretly, protect chairs from disputes about the draft before the room, and protect entities from giving authority to a moving entity. The AFPUB-2019-V4-003 appeal demonstrates the institutional cost when version history itself becomes contested.
Archiving closed the file but not the policy problem
AFRINIC's public archive marks draft 4 expired and archived. This is a terminal administrative status: the proposal is no longer active in the ordinary queue. It preserves the document and history for reference rather than allowing an unresolved draft to remain permanently current.
Archival closure is necessary. Policy communities have finite attention, and stale proposals can confuse entities about applicable text. Authors may leave, circumstances may change and a newer proposal may address the same subject more effectively. A clear archive prevents zombie authority.
Yet “expired” does not explain why the proposal failed to become policy. Did it lack substantive consensus? Did appeal unsettle the determination? Did time limits run out? Did another proposal supersede it? Did compatibility remain unresolved? Different causes carry different lessons.
A terminal report should answer those questions and identify which parts of the problem remain open. If another transfer proposal continues the work, map concepts rather than implying a clean replacement. If no active proposal exists, state the operational baseline that remains.
The transfer demand did not vanish because draft 4 entered the archive. Scarcity, market activity and inter-regional compatibility continued. Archive status ended one vehicle, not the underlying governance responsibility. A durable institution learns from the vehicle's failure before starting another.
A rival proposal can fragment the constituency
The AFRINIC records around this period refer to more than one transfer proposal, including alternatives with different scope. Parallel drafts can be healthy. They reveal competing principles and prevent one author from controlling the solution. They can also fragment attention and make consensus harder to interpret.
Entities may support the objective while splitting among texts. An objection to draft 4 may reflect preference for another proposal rather than opposition to transfers. Authors may withdraw conditionally, merge language or continue in parallel. Chairs need to distinguish policy disagreement from vehicle selection.
A comparative matrix should place proposals side by side: intra- or inter-regional scope, reciprocity, needs test, legacy status, source eligibility, recipient eligibility, holding period and registry duties. The community can then decide principles before choosing text.
If one proposal is archived because another supersedes it, the decision should be explicit. If both fail independently, the record should preserve why. Quiet around one draft cannot be counted as consensus for its rival unless entities affirm that movement.
Parallel work also raises timing concerns. External compatibility feedback on one version may apply to another, but only if wording is equivalent. Staff should not transfer conclusions without showing the match. The proliferation of identifiers makes a shared issue ledger even more important.
The goal is not to eliminate competition among ideas. It is to prevent competition from turning the constituency into several shrinking rooms whose silence is misread as closure.
External events can distort deliberative time
The 2019-2022 period included operational, legal and institutional pressures beyond this proposal. Policy volunteers and staff did not deliberate in a vacuum. Meeting formats changed, organisational attention was contested and resource governance carried increasing stakes. These conditions can slow response or intensify urgency.
External pressure does not automatically excuse delay or invalidate consensus. It changes the evidence needed. Chairs should state when a meeting disruption, staff constraint or legal issue affects the timetable. Lost participation windows should be restored. Emergency compression should narrow the decision to what is necessary.
Long proposals are vulnerable to context drift. A text developed under one scarcity assumption may be assessed later under a changed market or counterpart policy. Before reviving an old draft, staff should refresh implementation analysis and external compatibility. Prior consensus on factual premises may no longer hold.
At the same time, crisis can make entities accept weak closure simply to move on. The institution should resist using fatigue or urgency to transform a disputed proposal into an inherited baseline. If a decision cannot be supported, archive the text with reasons and open a cleaner question when capacity returns.
The phrase outlived its forum captures this temporal mismatch. The draft identity remained, while the meeting, entities and external circumstances that produced earlier judgments had passed. Authority needed renewal, not assumption.
Delay distributed economic advantage
Transfer policy delay is not neutral. Organisations with surplus registered IPv4 space, organisations needing addresses, brokers, hosting providers and networks able to obtain resources through other regions face different options while AFRINIC rules remain uncertain. Existing holders may benefit from scarcity; constrained entrants may bear cost; unrecognised transactions may move outside transparent registry channels.
This does not mean speed should override safeguards. A flawed transfer policy could facilitate harmful outflow, inconsistent registration or speculative concentration. The point is that “continue discussion” is itself a policy choice with distributional effects.
Every extension should therefore state the expected benefit of more time and the cost of delay. What evidence is missing? Who will obtain it? When will the next decision occur? If the answer is simply “more discussion,” the process has not justified the burden.
Staff can publish aggregate evidence about transfer inquiries, unmet requests and compatibility questions without exposing confidential transactions. This helps the community weigh urgency. Authors and objectors should also state who bears the risk of their preferred timing.
A time-bounded return can be legitimate: revise one clause, seek counterpart confirmation, publish analysis and call again by a fixed date. An indefinite return externalises cost onto people who may not participate in the forum. Accountability requires that the process see them.
Policy fatigue should trigger a different mode
After several versions and repeated debate, continuing the same form of discussion may produce diminishing returns. Entities restate positions, new readers struggle with history and chairs face the same disputed signals. The institution needs a rule for changing mode before fatigue becomes the deciding force.
One option is a structured issues conference. Freeze the draft, identify no more than a handful of unresolved questions, commission neutral staff analysis and hear proponents and objectors against the same evidence. Publish a disposition for each issue. Then either produce a final text for review or close the proposal.
Another is severance. If the community agrees on intra-regional mechanics but not inter-regional outflow, adopt or develop the separable part while continuing the harder question. Severance works only if the parts are operationally independent and entities review the assembled effect.
A third is terminal reset. Archive the draft without prejudice and require any successor to begin with a fresh problem statement, compatibility matrix and lessons report. Prior contributions remain evidence but no status carries over automatically.
The right choice depends on the proposal. What matters is recognising fatigue as a governance condition. Continuing to invite unrestricted comments after the forum has stopped learning is not openness; it is an abdication of decision.
A law of proposal ageing would improve finality
AFRINIC could adopt a public ageing standard for proposals. At defined intervals, perhaps six and twelve months, chairs would publish a status review: exact version, participation, unresolved issues, compatibility checks, author engagement, next decision and whether circumstances changed. A proposal could not remain active without a reasoned extension.
Ageing should not impose automatic rejection. Some global policies genuinely require years. The review would force a choice among active development, structured resolution, merger, withdrawal, rejection or archive. Each status would have a consequence and date.
Material revision would reset only the review needed for changed clauses, not erase the age of the underlying question. This prevents authors from avoiding closure through serial numbering. Conversely, institutional delay should not count against an author where staff analysis or appeal remains outstanding; the record should assign responsibility.
Age data should be reported across proposals. If certain subjects or authors consistently wait longer, the Board and community can examine capacity, conflict and criteria. Median time alone is limited public evidence; long-tail cases such as draft 4 deserve individual explanation.
A law of ageing makes time visible as governing power. It protects the community from stale drafts and protects difficult proposals from quiet neglect. Most importantly, it prevents expiry from being the first moment at which entities learn that the institution has stopped deciding.
Appeals need a remedy clock
Formal appeal can improve a contested consensus finding, but an unbounded appeal adds another layer of uncertainty. The review body should publish acceptance, questions, submissions, hearing dates, expected decision and any extension. The disputed text and status should remain frozen.
The remedy must be clear. If the error concerns incomplete revision history, require corrected disclosure and renewed review. If chairs failed to address a substantial objection, return the defined issue. If the appeal lacks merit, uphold the determination with reasons. If circumstances make the proposal obsolete, say so rather than deciding an academic question while the active process moves elsewhere.
Parties should know whether policy development continues during appeal. Parallel revision can cure defects but can also make the appeal target disappear. A clean rule can pause material advancement while allowing staff research and successor planning.
The Appeal Committee should have access to complete public records and any necessary confidential material under appropriate safeguards. Its decision should identify the standard of review: procedural compliance, reasonableness of consensus judgment or de novo policy assessment. Those are different roles.
Appeal success rates are not the measure. Quality appears in timeliness, specificity and whether the remedy restores an auditable path. AFPUB-2019-V4-003 shows why a policy community needs review; it also shows that review must lead somewhere.
The Board should oversee process health without choosing clauses
AFRINIC's Board has responsibilities for the organisation and for ensuring a functioning policy environment, while bottom-up development belongs to the community process. The Board should monitor ageing, appeal delay, staff capacity and compliance with published rules. It should not privately select its preferred transfer conditions.
Oversight can take the form of aggregate reports and procedural directions: require timely minutes, complete revision tables, compatibility assessments and reasoned terminal statuses. If the Board identifies legal or fiduciary risk in a draft, it should publish the concern at an appropriate level and return it through the process.
The distinction is important where valuable IPv4 resources are involved. Trustees may face claims from members, courts or counterpart institutions. They need reliable policy and cannot ignore operational risk. But using corporate authority to rewrite a community proposal would make consensus ceremonial.
The Board can also ensure that appeal bodies are staffed and independent, that chairs receive support and that public archives remain available. These institutional conditions may matter more to finality than intervention in any clause.
After a case like draft 4, the Board should commission a process review focused on time and evidence, not the merits of transfer policy. Which stage failed to resolve the issue? When did compatibility feedback arrive? Why did version history become disputed? Did appeal complete within a reasonable period? The answers improve future governance without deciding what transfer rule the community should prefer.
An auditable transfer-policy path
A better path for a future inter-regional transfer proposal would begin with a problem statement and policy-options matrix. The community would first discuss principles: permitted directions, needs basis, legacy treatment, reciprocity and registry responsibility. Chairs would record findings issue by issue.
Authors would then draft exact language against those findings. Staff would produce operational journeys and a legal summary. Counterpart registries would receive the frozen text and return version-specific compatibility assessments. Every material change would map to a concern and receive proportionate review.
Before a consensus call, chairs would publish the entity and objection record. The call would identify whether it covers principles, clauses or the full instrument. A last review period would test the exact text. Appeals would have fixed questions and remedy clocks. Board ratification or implementation review would remain distinct from the community finding.
If the proposal failed, the terminal report would say why: limited public evidence support, unresolved substantial objection, external incompatibility, author withdrawal, supersession or expiry after a reasoned ageing review. The archive would preserve reusable evidence for a successor.
This structure cannot guarantee agreement. It can guarantee that disagreement ends in a legible state. That is the missing public good exposed by a proposal that persists across revisions and forums without an accepted conclusion.
What draft 4 can and cannot establish
The public record establishes that a Resource Transfer Policy entered AFRINIC's process in 2019, reached at least four versions, encountered counterpart compatibility questions, received a contested consensus determination, entered formal appeal and was eventually marked expired and archived. These are verifiable institutional events.
The record also establishes that material policy choices changed or were disputed across versions, including needs assessment, legacy status and registry roles. The appeal document preserves the challenger's account and textual comparisons. It is evidence of controversy, not automatic proof that every allegation was upheld.
The record does not justify claims that authors, chairs, objectors, staff or trustees acted from improper motives. Delay can arise from complexity, volunteer capacity, external dependency and procedural uncertainty. Nor does archive status prove the policy idea lacked support. It identifies the fate of this proposal.
These boundaries matter because AFRINIC's governance has attracted intense conflict. A responsible analysis should not turn one policy dispute into a character judgment or borrow unrelated controversies. The case is sufficiently important on its own: a valuable resource rule moved through a process that struggled to produce accepted closure.
Precision strengthens criticism. The problem is not that disagreement lasted. The problem is that repeated activity made it difficult to tell which objections remained substantial, which text held authority and what event would finally settle the question.
Diligence needs a stopping rule
Careful policy takes time. Inter-regional IPv4 transfers require coordination across legal entities, resource histories and incompatible rules. A quick majority could have produced a mechanism no counterpart could execute. Revisions and appeal can therefore be signs of institutional seriousness.
Seriousness becomes stasis when no actor is responsible for converting evidence into a bounded next step. More comments do not always improve a draft. More versions do not always narrow disagreement. More time does not turn silence into consensus. At some point, the institution must decide, sever, reset or archive with reasons.
A stopping rule should not privilege the status quo invisibly. If unresolved objection prevents adoption, chairs must identify it and explain its weight. If the objection has been answered, they must be willing to close despite continuing preference. If external compatibility is impossible, say so. If the community lacks participation, conduct targeted consultation and then decide.
The rule protects objectors as well as proponents. A clear rejection preserves the concern and permits a successor proposal. Indefinite limbo exhausts everyone and allows informal practices to fill the gap without public authority.
AFPUB-2019-V4-003-DRAFT04 outlived its forum because the proposal's identity persisted beyond the point at which ordinary meeting iteration could supply accepted finality. The lesson is not to hurry the next transfer policy. It is to give diligence a destination.
Archive should become institutional memory
AFRINIC's archive preserves draft 4 rather than erasing it. That is a foundation for learning. The organisation can add a closing memorandum connecting versions, external feedback, consensus determination, appeal and terminal status. Such a memorandum would not relitigate merits; it would make the institutional path intelligible.
Future authors could see which clauses failed compatibility and which drafting choices attracted repeated concern. Chairs could identify where summaries were limited public evidence. Staff could reuse operational analysis. Appeal bodies could point to clearer standards. Entities would not have to reconstruct the episode from scattered pages and messages.
The memorandum should preserve uncertainty and disagreement. It can state that an appellant alleged an omitted revision and record the review outcome if available, rather than presenting allegation as fact. It can identify unresolved issues without assigning motive. It can distinguish this proposal from successors with similar titles.
Institutional memory is especially important in number policy because scarcity questions recur. A later proposal may use different language while confronting the same needs, legacy and reciprocity trade-offs. An archive that records only “expired” forces the community to repeat the argument.
The final measure of closure is not whether a page moved into an archive menu. It is whether the institution retained enough reasoned knowledge to make the next decision better.
A proposal can end without the question being defeated
Policy communities sometimes resist terminal decisions because authors fear that rejection will be read as rejection of the problem. That fear encourages endless revision. The institution should separate vehicle from issue. AFPUB-2019-V4-003 could expire while the need for inter-regional transfer policy remained real.
A terminal report can say that the draft did not secure an accepted path because of specified unresolved issues, while inviting a successor built on the lessons. It can preserve agreed principles without carrying over disputed text. It can also identify the current policy baseline so operators know what remains possible.
This distinction reduces the stakes of each version. Authors can withdraw a flawed text without abandoning their objective. Objectors can support continued work without endorsing the current clauses. Chairs can close stale files without being accused of silencing the subject.
The same principle applies to appeal. Upholding a procedural challenge does not decide transfer policy. Rejecting an appeal does not prove the draft wise. Each institution should make only the claim within its authority.
Draft 4's afterlife is most useful when understood this way. It is not a monument to failure or a policy waiting to be revived unchanged. It is a documented attempt whose unresolved journey reveals design requirements for whatever comes next.
Finality is part of bottom-up legitimacy
Bottom-up governance is often defined by openness: anyone may join the list, proposals come from the community, meetings are public and consensus replaces command. Openness without finality can still concentrate power. People with the stamina to remain through years of revision gain influence, while small operators and occasional experts leave. Staff administer the status quo during uncertainty. Chairs control when discussion ends.
Finality does not mean every entity agrees. It means the institution has a known method for turning reasons into an authorised state: adopted, rejected, returned on defined issues, withdrawn, superseded or archived with explanation. Appeal reviews the state without dissolving it into endless reconsideration.
For AFRINIC transfer policy, finality had additional public value. Counterpart registries needed to know whether a text could support reciprocal operation. Members needed to know what transactions the registry would recognise. Market entities needed accurate registration rules. Ambiguous status imposed costs outside the policy forum.
The governance reform is therefore procedural but not cosmetic: clause-level consensus records, version accountability, compatibility checkpoints, ageing reviews, fatigue triggers, appeal clocks and explanatory terminal reports. Together they make participation worth the effort because contributions lead to a visible disposition.
AFPUB-2019-V4-003-DRAFT04 shows that a proposal can receive enormous attention and still leave authority unclear. The answer is not less discussion. It is discussion designed to end in a reasoned state.
The forum must know when its work is complete
The Resource Transfer Policy case began with a legitimate and difficult problem. It asked AFRINIC's community to connect African registry principles to an inter-regional system under IPv4 scarcity. Four drafts, counterpart feedback and appeal show that entities took the consequences seriously.
The same history exposes the limit of repeated iteration. Material choices moved across versions. External compatibility challenged local wording. A consensus declaration failed to produce accepted closure. Appeal extended the institutional life of the dispute. Archive status eventually ended the active file without, by itself, explaining every lesson.
A forum completes its work when it can state the exact text, evidence, remaining objections, authority and consequence of its decision. Completion may be adoption. It may also be a reasoned no. What it cannot be is perpetual invitation after the process has stopped generating new understanding.
Future AFRINIC proposals should treat time as part of policy accountability. Each return must name the missing evidence and next date. Each revision must expose changed power. Each external dependency must be tested against a frozen text. Each appeal must have a remedy clock. Each archive entry must explain what ended and what remains.
Those controls would not make inter-regional transfers uncontroversial. They would make controversy governable. Draft 4's legacy should be that distinction: diligence is the disciplined pursuit of a decision; stasis is activity without a credible stopping rule. A bottom-up institution owes its entities the first and must be willing to name the second.