Summary
- AFRINIC became the current administrative home for African number-resource records after a phased 2005 transition from APNIC, ARIN and RIPE NCC, but current custody is not the same thing as proof of original allocation terms, holder identity or title.
- The fixed public materials establish service continuity: co-evaluation from September 2004, AFRINIC's primary service role on 21 February 2005, continued second-opinion review, March 2005 delegated-stat snapshots and 8 April 2005 recognition.
- The same materials do not enumerate a resource-level conveyance schedule, field-by-field database reconciliation, executed service-agreement transfer, liability assignment, known erroneous row or decided correction appeal.
- A responsible legacy-record audit must separate service responsibility, registry data, supporting applications, agreements, contact succession, correction history and appeal reasoning before assigning title-like weight to an inherited entry.
A row at the cutover
Imagine one disputed African IPv4 allocation as it stood during the February 2005 handover. The example does not assume the holder, prefix size, predecessor registry, date, contact name, contract terms or defect. It is useful precisely because those facts are withheld. A party later says the row proves a durable right. Another says the name in the record was stale, the address block had moved through corporate succession, or the listed contact had no authority. AFRINIC, as the current registry, can point to the live administrative record. That is a necessary starting point. It is not enough to settle every historical question.
The evidentiary stack for that row has several layers. The first is a delegated-stat row: registry, resource type, country code, start value, count, date and status as preserved in a dated statistics file. The second is the full registry record, including contact handles, organisation fields, remarks, update history and any successor record not visible in the statistics file. The third is the supporting application or approval material that explains why the resource was issued. The fourth is the service agreement or membership relation that governed the holder's dealings with the relevant registry.
The fifth is the contact and succession history that links the recorded party to a real company, institution or person over time. The sixth is any correction request, objection, escalation or appeal that shows how a later dispute was handled.
Those layers answer different questions. A delegated-stat entry can help prove that a summary row existed on a certain date. It cannot prove beneficial control. A full registry record can show administrative fields. It cannot prove that every underlying document was complete. An application can explain original justification. It may not show later succession. An agreement can identify service terms. It may not define property title. A contact record can support operational authority. It may not survive a merger, liquidation or staff departure. A correction or appeal file can show institutional reasoning.
It may not exist for a row that no one challenged at the time.
The transitional problem is that AFRINIC inherited many such records from a system split among APNIC, ARIN and RIPE NCC. The public record shows that the service role moved to AFRINIC. It does not show, row by row, which original allocation packet, contract file, liability trail and correction history accompanied each entry. This distinction does not undermine AFRINIC's current authority to administer the registry. It defines what additional proof is needed when an inherited row is asked to carry title-like weight.
The transition as a service event
Before AFRINIC became the recognised regional registry for Africa, parts of the continent were served by three incumbent registries: APNIC, ARIN and RIPE NCC. That arrangement was administratively awkward. A single African service region could not mature cleanly if its registry support remained scattered across three institutions with their own systems, historical files and operational habits. The creation of AFRINIC addressed that fragmentation by giving the region a dedicated service institution.
The key public materials describe a staged transfer rather than an instantaneous proof of every underlying fact. The updated AFRINIC application described training, co-evaluation, second-opinion arrangements, migration of existing RIR service agreements and distributed operating functions. The NRO progress update recorded that AFRINIC systems took the primary role on 21 February 2005 while incumbent registries continued second-opinion review. ICANN's public notice later said that APNIC, ARIN and RIPE NCC had transferred all RIR services to AFRINIC on that date and that AFRINIC was operating as a functional RIR.
That is strong evidence of service continuity. It tells operators where requests would be handled, which institution had the primary operational desk, and how external review reduced cutover risk. It also shows that the transition was not a mere aspiration. AFRINIC had moved from preparation into active administration, with incumbent review still present until final recognition.
The same record is not a resource-level conveyance schedule. It does not list each block, each holder, each agreement, each predecessor file, each unresolved identity conflict or each allocation packet. It does not say whether every contact field was reconciled against corporate succession records. It does not identify every supporting document that moved, every document that stayed with a predecessor registry, or every liability that might have accompanied service history. The phrase "all RIR services" therefore has to be read as a continuity statement. It means service responsibility moved.
It does not mean every evidentiary layer for every resource was exposed to public inspection.
This distinction is practical, not semantic. If the issue is whether AFRINIC was the correct registry to receive new African service requests after 21 February 2005, the transition record is directly relevant. If the issue is whether one inherited row proves that a particular company held a transferable title on a particular historical theory, the transition record is only the first layer. The second inquiry needs resource-specific evidence.
Four dates with different meanings
The chronology matters because each date has a different evidentiary function. September 2004 marks the beginning of co-evaluation and provisional preparation. It shows that the new registry was not entering 2005 without prior review and training. It is not the date on which every African resource record became a newly created AFRINIC entry.
21 February 2005 is the primary-service transfer date. The available record says AFRINIC systems assumed the primary role while the incumbent registries continued second-opinion review pending recognition. That date matters for service responsibility. It does not by itself identify which supporting records, agreements, notices, correction files or liabilities moved with each resource.
March 2005 matters for two different reasons. AFRINIC's updated application supplied a structured institutional account of readiness, training, service migration and records capability. The IANA delegated-stat mirror for AFRINIC in 2005 also preserves daily files beginning in March, giving later reviewers a reproducible way to compare summary rows after the service transition began. Neither March material is a full registry database. The application is an applicant account. The statistics are summary files.
8 April 2005 is the recognition date in the sequence required here. It marks the final institutional act by which AFRINIC was recognised as the regional registry for the African region. Recognition explains why AFRINIC became the central registry authority. It does not retroactively make every inherited field an original AFRINIC-created fact.
The IANA assessment that preceded recognition is also important. It stated that incumbent RIRs had served the region, that joint communications went to their members, that service agreements could be migrated and that AFRINIC had record-keeping capability. If discussing its internal chronology, the September 2005 reference must be corrected to the proven September 2004 provisional step. The assessment supports aggregate readiness. It does not publish a resource-by-resource reconciliation file.
Keeping the dates separate prevents two errors. One error treats recognition as if it created all prior facts. It did not. The other treats inherited records as suspect merely because they were inherited. That is also wrong. The transition created a current administrative home for records made under three prior systems. The present question is how much proof a given inherited row can bear without its surrounding file.
The three-predecessor inheritance problem
AFRINIC's inheritance was more complex than a simple office relocation. It was not one predecessor registry handing a single coherent African archive to one successor. It involved three incumbent RIRs that had served different parts of the region under their own practices. That structure raises a governance issue: centralisation solved a service problem while creating a later evidence problem for any row whose origin, holder identity or supporting file becomes disputed.
APNIC, ARIN and RIPE NCC each had relationships with networks in parts of Africa before the transition. Those relationships could include service agreements, allocation records, administrative contacts, supporting applications, update practices, notices and local interpretations of policy. The public materials say service responsibility moved. They do not give a full inventory of what each predecessor held, how each item was classified, how each document was matched to a resource, or how inconsistencies across predecessor records were resolved.
A three-source reconciliation would therefore need to ask five separate questions. First, who handled service responsibility for the resource before the handover and after it? Second, what agreement or membership relation governed the resource holder before and after migration? Third, which registry fields and supporting documents were transferred, copied, cross-referenced or retained by the predecessor registry? Fourth, what liabilities, limitations or historical caveats accompanied the record? Fifth, what route existed for a holder to entity, correct or appeal if the inherited entry was wrong?
The available materials answer the first question at the aggregate level. They show a service transfer to AFRINIC on 21 February 2005 and continued second-opinion review during the transition. They provide some support for the second question by referring to migration of existing RIR service agreements. They provide only limited support for the third by noting record-keeping capability and by preserving later delegated-stat files. They do not answer the fourth or fifth at resource level.
That uneven evidence matters when records are used in disputes. A current registry entry can be authoritative for administration without being conclusive for historical title. A court, arbitrator, registry appeal panel or governance reviewer would need to know whether the row came from APNIC, ARIN or RIPE NCC, what documents supported it, and whether the holder had notice and a chance to correct problems. The public transition record alone does not supply those details.
The point is not that inherited rows are unreliable. The point is that reliability has to be shown through layers. A row that was stable across predecessor files, application packets, service agreements, contact history and later AFRINIC updates can be strong evidence. A row that appears only as a summary statistic, with missing support and disputed identity, should bear less weight. The transition record cannot tell those two rows apart unless it is joined with resource-specific material.
Service responsibility, agreements, records, documents and liabilities
The phrase "transferred all RIR services" does valuable work, but it should not be stretched beyond its function. In registry operations, service responsibility means the institution that receives requests, applies policy, maintains records, communicates with holders and coordinates with the global registry system. Moving that responsibility to AFRINIC was necessary for regional continuity. It does not follow that every contract term, evidence file and liability issue became public or settled.
Service responsibility is the cleanest layer. The transition materials identify 21 February 2005 as the date on which AFRINIC became primary for the African service function, subject to continued second-opinion review. This gives a clear administrative answer: after that date, AFRINIC was no longer merely preparing. It was operating the desk.
Agreements are a different layer. The updated application refers to migration of existing RIR service agreements. The IANA assessment says service agreements could be migrated. Those statements matter, but they are aggregate descriptions. They do not attach executed schedules naming every party, resource, predecessor registry, governing term, consent notice or exception. For a disputed row, the agreement question cannot be settled by the aggregate statement alone.
Records are the third layer. AFRINIC had to maintain a current registry, and the recognition assessment treated record-keeping capability as part of readiness. Still, record-keeping capability is not the same as field-level reconciliation. A current database can preserve inherited data accurately, carry inherited errors forward, or contain a mixture of confirmed and unconfirmed fields. The public documents do not measure those categories.
Documents are the fourth layer. Original applications, approvals, correspondence, contact changes, corporate name changes and supporting technical justifications may be separate from summary registry fields. A transition can maintain service while not placing every underlying document into an easily reviewable public archive. For title-like reliance, document custody matters because it explains why the row exists and who had authority to ask for it.
Liabilities are the fifth layer. The public materials do not identify what liabilities, if any, moved from predecessor registries to AFRINIC for historical acts, stale contacts, incorrect fields or missing supporting files. That absence should not be converted into a claim that liabilities did or did not move. It is simply an unmeasured part of the record.
Separating these layers prevents overclaiming. AFRINIC could be the correct current administrator even if a particular agreement file was incomplete. An inherited contact could be correct even if the public transition file does not prove it. A disputed title claim could fail even if the row appears in a current registry. The question in each case is which layer is being asked to do which job.
What delegated-stat files can prove
Delegated-stat archives are among the most useful public tools for testing the transition because they are dated, reproducible and structured. The March 2005 files in the IANA mirror provide snapshots that can be compared with predecessor files and later AFRINIC records. A reviewer can identify whether a resource appears in a summary row, whether its date and status fit the expected pattern, and whether later files show continuity or change.
The RIR statistics exchange format explains why those files are useful and limited. The files are built from defined fields such as registry, country, resource type, starting value, count, date and status. That structure allows comparison across time. It is a public way to test whether a row existed in a statistics archive and how it was summarized.
But a delegated-stat file is not a full WHOIS history. It does not include every contact handle, email address, organisation record, remark, internal note, application packet, agreement, identity proof, objection, update event or appeal. It is not an ownership register. It is not a contract file. It is not a statement that the party listed elsewhere had beneficial control. It is not proof that a predecessor registry verified every later succession claim.
That does not make the archive weak. It makes it precise. A delegated-stat row is strong evidence for the existence of a summary state at a date. It is weaker evidence for why the row existed, who controlled the resource, what agreement applied, whether the holder consented to migration, or whether later corrections were justified. The responsible use is to let the file prove what its format can carry and then seek other evidence for questions outside that format.
This is especially important for inherited African records because the transition crossed institutional boundaries. If a row appears in a March AFRINIC delegated-stat file, that may confirm that AFRINIC's public statistics included the resource after the handover. To evaluate provenance, one would still compare the row with APNIC, ARIN or RIPE NCC predecessor files, identify the matching holder record, inspect supporting documents and check later AFRINIC changes. The archive opens the audit. It does not finish it.
Delegated-stat archives also reduce the risk of pure memory-based disputes. They give later reviewers a common reference point. A party cannot simply assert that a row never existed if a dated file shows it. Conversely, a party cannot claim that the row proves everything if the format never carried the disputed fact. The public archive is valuable because it disciplines both exaggeration and denial.
When inherited evidence becomes stronger
An inherited row becomes stronger when separate records converge. Suppose a March 2005 AFRINIC statistics file shows a resource, a predecessor registry file shows the same resource before the handover, the full registry record carries a consistent holder name, the supporting application identifies the same institution, and the service agreement ties that institution to the relevant registry relation. In that situation, the current row is no longer standing alone. It is part of an evidentiary chain.
Convergence does not require every field to be identical. Historical records often contain changes in spelling, contact format, postal addresses or organisation labels. The issue is whether the differences are explained by ordinary updates or by unresolved identity problems. A new contact for the same institution may be routine. A new entity with no documented succession is different. A date change that reflects a known registry update is different from a date change that conflicts with the allocation file.
The strongest evidence is not simply more paper. It is alignment across functions. A statistics file shows public summary state. A full registry record shows administrative details. An application shows original justification. An agreement shows the service relation. Contact and succession documents show authority. Correction records show how the registry treated later disputes. Each layer reduces a different uncertainty.
This approach also helps avoid unfair exclusion. Legacy records should not fail merely because one old file is missing. The correct question is whether the remaining evidence can still explain origin, identity, authority and continuity. A row supported by multiple independent layers may be credible even if one minor document is unavailable. A row supported only by current custody may still be usable for ordinary service, but it should not carry the same weight in a title-like dispute.
The three-predecessor setting makes this convergence test especially important. A row inherited from APNIC may have different surrounding records from one inherited from ARIN or RIPE NCC. A uniform AFRINIC database view after 2005 could mask different original file histories. A serious audit should therefore preserve predecessor context rather than flatten every inherited entry into one new category.
Predecessor context also matters for notice. If incumbent registries sent joint communications to their members, as the IANA assessment indicates at the aggregate level, a disputed case should ask whether the affected holder fell within the notice path, what the communication said, and whether it gave a meaningful chance to act. Aggregate notice supports institutional transition. Case-level notice supports reliance against a particular holder.
Consent is another layer that should not be assumed from silence. Service migration may have been necessary for continuity, and it may have been broadly supported. That does not prove every holder consented to every later interpretation of its resource status. Where a claim depends on consent, the evidence should identify the agreement, notice, acceptance conduct or policy basis that supplies it.
The same discipline applies to title-like claims made by holders. A holder cannot rely on registry history selectively. If it wants the current row to prove a durable right, it should also be prepared to show the predecessor allocation basis, agreement terms, identity continuity and absence of a contrary correction record. A record can be authoritative for service without giving either the registry or the holder a shortcut around the underlying file.
This symmetry matters. The custody distinction should not be used only against AFRINIC. It also prevents resource holders, purchasers or successors from treating a convenient row as a complete chain of rights. The public interest lies in matching the claim to the evidence. Administration needs one current registry. Historical rights need the documents that created and preserved them.
What a reasoned correction route must show
The public transition materials do not provide a decided correction appeal for an inherited African row. That absence leaves a governance question: what would a credible correction route need to show if a legacy holder challenged an inherited entry? The answer begins with intake. The registry should be able to identify the disputed field, the claimant, the asserted authority, the requested change and the evidence offered.
The second element is preservation. When a legacy row is challenged, the registry should preserve the current state and the known predecessor state before making changes. That allows later reviewers to understand what was altered and why. Preservation is especially important when a correction could affect contact authority, service access, transfer claims or public statistics.
The third element is source separation. A correction decision should not treat all evidence as equal. A statistics row proves summary state. A full registry record proves administrative fields. An agreement proves service terms. Corporate documents may prove succession. Internal correspondence may explain why a past update occurred. The decision should state which evidence was used for which fact.
The fourth element is burden allocation. If a party seeks a routine contact update, the evidence burden may be lighter than if it seeks recognition as successor holder for a legacy block. If the registry proposes to change a contested holder field, it should have a reasoned basis. If a claimant seeks title-like recognition, the claimant should provide the chain of authority. The heavier the consequence, the more complete the file should be.
The fifth element is notice to affected parties. A correction that alters identity or control should not be made on a one-sided record when another plausible holder or contact exists. Notice does not mean that every old email will work or that every dormant entity can be reached. It means the registry should record reasonable attempts and explain why action is justified despite limits.
The sixth element is a reasoned outcome. A decision should state whether the registry accepted, rejected or partially accepted the correction, and why. It should distinguish administrative correction from recognition of broader rights. For example, updating a contact field because a company changed staff is not the same as confirming that the company owns the resource as property.
The seventh element is review. A correction route has more legitimacy if a disappointed party can seek reconsideration or appeal under known rules. Review does not guarantee a different result. It creates a record that the registry's inherited custody was tested through reasons rather than merely asserted through current control.
The eighth element is public-safe documentation. Legacy disputes may contain private information, contracts or sensitive security data. A registry need not publish raw files to show legitimacy. It can preserve the file internally and, where appropriate, publish anonymised statistics or reason categories. That would help later users understand whether inherited-error risk is isolated or systemic without exposing private documents.
Such a correction route would serve both continuity and fairness. It would allow AFRINIC to maintain one authoritative registry while giving disputed rows a path toward evidence-based resolution. It would also create the missing denominator over time: how many inherited records were challenged, what kinds of defects were alleged, how many were corrected, and how many decisions survived review.
Without that kind of route, inherited custody carries two opposite risks. The registry may over-rely on a row because it is current. A claimant may over-rely on the same row because it is convenient. A reasoned correction system prevents both. It treats the current record as the administrative base and the supporting file as the evidence needed for higher-stakes conclusions.
The strongest defence of centralisation
The strongest defence of AFRINIC's inherited custody is not convenience. It is the need to avoid indefinite fragmentation. If African resource records had remained scattered across APNIC, ARIN and RIPE NCC, operators would have faced inconsistent service routes, uneven update practices and greater uncertainty about which institution could act. A single regional registry reduced that institutional confusion.
Centralisation also supported accountability. With one registry responsible for the region, holders could know where to submit requests and where to seek administrative answers. Coordination with the global registry system became cleaner. Policy discussion could be attached to a regional institution rather than split across three external desks. That was not a minor benefit for a continent-wide service region.
The phased method strengthened the defence. AFRINIC did not simply receive a database and operate without review. The transition materials describe training, co-evaluation and continued second-opinion review. That arrangement reduced the risk of a reckless cutover. It allowed AFRINIC staff to apply inherited administrative practice while the incumbent registries still reviewed decisions during the period before final recognition.
The defence also includes practical necessity. Registry systems need a current authoritative record. Network operators cannot wait for perfect historical files before every update. If a holder needs a correction, a contact update or a service action, the registry must have an operational view from which to proceed. A system that treats every inherited row as unusable until the full archive is reconstructed would fail the continuity purpose of the transition.
For that reason, inherited custody should be respected as administrative authority. AFRINIC's current record can be the operative record for registry service. It can guide updates, contact handling, policy application and public statistics. It can be presumed usable for ordinary administration unless a dispute or risk marker gives reason to inspect deeper layers.
Yet the same defence supports better evidence, not less. Because one registry became the central authority, any inherited error could gain greater practical force. A stale contact in a scattered predecessor file might have had limited reach. The same contact in a central successor registry could shape later service decisions, transfer debates, compliance action or claims about the identity of the holder. Centralisation solved fragmentation and increased the cost of unresolved defects.
The danger of self-proving custody
Self-proving custody is the mistake of treating the current record keeper's row as proof of every fact that led to the row. In ordinary administration, that shortcut is attractive. The current registry has to act. It cannot relitigate every old allocation each time a ticket arrives. But disputes over legacy identity, succession, title-like claims or contested authority require a more careful method.
The first risk is identity drift. A resource may have been issued to a company that later changed name, merged, dissolved, sold assets or delegated operations to another entity. A current registry row may preserve one label while the legal or operational reality changed. Without succession evidence, the row can identify an administrative contact without proving current holder identity.
The second risk is contact inflation. A person listed in a registry field may have had technical or administrative responsibility at one point. That does not mean the person could consent to migration, transfer title, bind a successor company or waive objections. Contact fields are operational tools. They are not automatically corporate authority instruments.
The third risk is contract uncertainty. The public materials refer to migration of service agreements, but they do not provide the executed schedules or terms for each resource. If a dispute depends on contractual rights, service limits or consent, the current registry row cannot replace the agreement file. It can point toward the relation; it cannot define all of it.
The fourth risk is error preservation. A successor registry may accurately preserve an inherited error. The accuracy of copying is not the same as accuracy of original fact. If a predecessor record contained a stale country code, outdated contact, incomplete organisation name or ambiguous holder field, faithful migration would carry the defect into the new system. The public transition materials do not measure how often that happened.
The fifth risk is appeal opacity. A disputed row gains legitimacy when the affected party can request correction, receive reasons and obtain review. The available materials do not include a durable resource-level appeal record for inherited entries. That does not prove that no route existed. It means title-like reliance should ask for the correction and appeal history before treating the row as conclusive.
These risks do not justify automatic distrust of AFRINIC's records. They justify calibrated reliance. A current registry record can be strong administrative evidence. It becomes stronger when supported by predecessor files, agreements, applications, notices, contact succession and reasoned correction decisions. It becomes weaker when it is isolated from those layers and asked to decide a question outside its format.
What the public materials do not enumerate
The updated application is an important transition document, but it is not a conveyance schedule. It does not enumerate every resource, every predecessor registry file, every holder notice, every contract term, every unresolved objection or every correction case. It describes institutional readiness and service migration at a high level.
The NRO progress update is important operational evidence, but it is not a database reconciliation report. It records the primary service role on 21 February 2005 and continued second-opinion review. It does not identify which fields were checked, which records were corrected, which files were missing or which predecessor decisions were later challenged.
ICANN's public comment notice is important because it records the service-transfer claim and functional operation. It is not a resource-level chain-of-custody file. Saying that all RIR services were transferred tells the public that AFRINIC was now serving the region. It does not publish a list of every record and supporting document that travelled with that service function.
The IANA assessment is important because it evaluates recognition criteria and confirms record-keeping capability. It is not a legal opinion on property title, beneficial control or successor liability. It does not expose a full support denominator, a complete case file, an agreement schedule or a correction docket for inherited resources.
The delegated-stat archive is important because it allows reproducible row-level comparison of summary data from March 2005 onward. It is not a full registry history. It does not contain the applications, agreements, contact authority proof or appeal reasoning needed to settle higher-stakes disputes.
These limits should be stated once and then used. If a later analysis asks whether AFRINIC became the operational registry, the public materials are strong. If it asks whether a particular inherited record proves title, those materials identify the need for additional files. The right conclusion depends on the question.
A practical audit protocol for inherited rows
A legacy-record audit should begin with the least controversial layer: public statistics. Identify the earliest AFRINIC delegated-stat row after the February transition and compare it with the relevant predecessor registry files where available. Record the date, status, resource type, start value and count. The purpose is not to prove title. It is to establish the public summary trail.
The second step is to identify the predecessor registry. Was the resource previously handled by APNIC, ARIN or RIPE NCC? The answer determines which historical practices, agreements and supporting files may matter. A three-predecessor transition cannot be audited as if every row came from one office under one file system.
The third step is to locate the full registry record and update history. Summary statistics should be joined with WHOIS-style records, organisation fields, contact handles, remarks, changes and later AFRINIC updates. A row that remained stable across full records and public statistics has a different evidentiary profile from a row whose contacts or organisation fields changed around the handover.
The fourth step is to locate the original application or allocation support. This may include the network justification, institutional identity, approval date and conditions used by the predecessor registry. If the original application is unavailable, the audit should mark that absence rather than fill the gap with confidence from the current row.
The fifth step is to inspect the service agreement or membership relation. The audit should identify the party, the date, the predecessor registry, the migration basis, any notice of service change and any term relevant to correction or dispute handling. Aggregate statements about agreement migration are not substitutes for the specific file when rights are disputed.
The sixth step is contact and succession review. The audit should connect the recorded holder or contact to real institutional continuity. Name changes, mergers, liquidations, sales, delegated operations and staff departures all matter. The question is not merely who appears in the field. It is whether the person or entity had authority for the action being claimed.
The seventh step is correction history. Search for requests, objections, registry responses and reasoned decisions. If the row was ever challenged, the reasoning matters. If it was not challenged, that fact has limited meaning unless the holder had notice and a practical route to entity.
The eighth step is appeal review. A decision gains evidentiary force when it can survive reasoned review. The public transition materials do not supply an appeal docket for inherited rows. A serious audit should therefore ask whether a correction route existed, whether it was used and what reasons were given.
The ninth step is a confidence classification. Some rows will be high confidence because statistics, full registry records, applications, agreements, contact succession and correction history all align. Others will be medium confidence because the public row is stable but supporting documents are incomplete. Others will be contested because identity, authority or agreement terms are uncertain. The classification should describe evidence, not reward the party that already appears in the current row.
This protocol is deliberately more demanding than ordinary registry administration. It should not be required for routine service. It is required when an inherited record is asked to settle historical title, disputed succession or high-impact control. The burden rises with the consequence of the claim.
The missing denominator
The most important unknown is the denominator of migration. The public materials do not state the total number of resource records moved from APNIC, ARIN and RIPE NCC into AFRINIC's operational custody. Without that number, no one can calculate the rate of corrected rows, conflicting contacts, missing agreements or appealed decisions.
A second missing denominator is the number of records corrected during or soon after the transition. If many rows required adjustment, the transition would still be valid, but reliance on unexamined inherited fields would need caution. If few rows required adjustment, that would support stronger confidence. The public record here does not measure either outcome.
A third missing denominator is the number of unresolved identity or succession cases. Legacy resources are especially sensitive to changes in corporate form and operational control. Knowing how many entries had unclear holder continuity would help distinguish isolated disputes from a broader class of risk. The available materials do not supply that measurement.
A fourth missing denominator is the number of agreements actually migrated and the number that required replacement, notice, consent or special handling. The aggregate record says migration was part of the plan. It does not list the executed results. For disputes that depend on agreement terms, the absence of a public schedule is material.
A fifth missing denominator is the number of objections and appeal outcomes. A registry can gain legitimacy by explaining why it accepted, rejected or modified inherited records. Published counts, anonymised reason categories or durable review records would help later users understand the correction system. The fixed materials do not provide those figures.
These gaps do not defeat the service transition. Infrastructure institutions often centralise first and refine evidentiary records over time. But the gaps do limit what can be inferred from current custody. A row can be good enough for present service and still not be enough for a high-stakes title claim without its supporting file.
The missing denominator also protects AFRINIC from unfair generalisation. Without measured incidence, critics cannot responsibly claim that inherited records were broadly erroneous. Supporters likewise cannot claim that inherited errors were negligible. The honest position is narrower: the public record proves aggregate service transition and supplies tools for comparison, while error incidence, correction rate and appeal outcomes remain unmeasured in the materials considered here.
The boundary between administration and title-like weight
Registry administration requires a current authoritative record. AFRINIC had to operate from a ledger after 21 February 2005. New requests, updates, contacts and policy applications needed a single administrative home. The transition materials show why that home existed and why it was needed.
Historical provenance asks a different question. It asks where the row came from, what documents supported it, who had authority, what terms applied, and how disputes were resolved. A current registry can preserve that history, but the current row is not automatically a complete history. It is one expression of the administrative state after migration.
Contractual rights are also distinct. A service agreement may govern obligations between a registry and a holder. It may explain fees, updates, contacts, confidentiality, compliance and remedies. The public transition materials' aggregate references to agreement migration cannot replace the specific terms of the relevant agreement when a dispute depends on them.
Property claims are the narrowest and most demanding. The available materials do not establish that every inherited resource row became proof of property title. They do not define title, transferability or beneficial control for each block. Any party making such a claim needs evidence beyond the fact that AFRINIC became the registry administrator.
This four-part distinction is the article's central answer. Current authoritative administration is real and necessary. Historical provenance is record-specific. Contractual rights depend on the agreement and governing instruments. Property claims require a separate basis. Confusing these categories gives too much or too little force to the 2005 transition.
AFRINIC's inherited pool should therefore be treated as operationally authoritative but evidentially layered. The registry's current custody can justify administration. It can start an inquiry. It can provide public statistics and records. It cannot, standing alone, make every inherited row self-proving evidence of original allocation terms, identity, succession, consent or title.
That is not a diminished account of the transition. It is the only account that fits the record. A scattered African service environment needed centralisation. AFRINIC received that role through training, co-evaluation, primary service transfer, second-opinion review, public notice, assessment and recognition. The same documents leave resource-level provenance to be proven through the files that actually carry it. Inherited custody keeps the registry functioning; evidence decides how much weight a contested row can bear.