RIPE NCC · RIR Watchdog

Initial Certification Policy in the RIPE NCC Service Region

Start here for the proposal-by-proposal record: what changed, why it mattered, who argued for what, how positions evolved, what was decided, and where the evidence still has gaps.

Source facts

Policy ID: 2008-08
RIR: RIPE NCC
Official status: Withdrawn
Normalized status: Withdrawn
Introduced: Unknown
Last source update: Oct 11, 2026
Coverage: Partial

Who actually shaped the debate?

3Total participants

2Discussion contributors

1Formal actors

3Discussion messages

1.8Effective participants

100%Top-five discussion share

1People covering 50%

2People covering 80%

BTW Analysis

What changed

The proposal evolved from a PA-only certification model with maintenance tied to membership or contractual status in version 1, to explicit dependence on the registered resource record in version 2. Version 3 framed the service as RPKI and added a member-in-good-standing condition. Version 4 broadened initial eligibility to IPv4 resources marked "ALLOCATED PA" or "ALLOCATED UNSPECIFIED" and IPv6 resources marked "ALLOCATED-BY-RIR", while continuing to defer PI, ERX and legacy resources. Across the versions, certificate validity and continuity remained tied to registration status. The proposal was ultimately withdrawn on 2011-07-26, so these changes describe proposed policy text rather than an adopted policy.

What the debate was really about

The central issue was who would control certificate eligibility and continuity, and which institutional conditions would serve as the gate. The drafts moved between contractual or membership status, the existence of a registered holding in the RIPE Database, and membership in good standing, while gradually expanding the resource categories covered. In practical terms, the proposal would have made certification depend on the RIPE NCC's registered-holder records and, in later versions, good-standing status. That is a structural shift toward institutional control over certification, but the supplied record does not establish realized operational harm or explain why the proposal was withdrawn.

Arguments that prevailed

Within the proposal texts, the supporting rationale for an initial, phased certification service prevailed during drafting: the drafts consistently focused first on comparatively simpler resource categories and deferred PI, ERX and legacy resources.
The successive drafts retained the principle that certificates should reflect the registered status of the underlying resource, and later versions retained or strengthened institutional eligibility conditions.
No argument can be identified as prevailing in the final policy decision, because the proposal's recorded outcome was withdrawal.

Arguments that did not prevail

The proposal's overall certification framework did not prevail as an adopted policy under 2008-08; the final recorded outcome was withdrawal.
Version 1's specific reliance on valid membership or another contractual relationship was not retained unchanged: version 2 shifted emphasis to the registered holding record, while version 3 restored an explicit good-standing condition.
The proposal's planned expansion from PA resources to the broader categories listed in version 4 did not become an adopted provision under this proposal.
The proposal documents record no opposing arguments, but the supplied evidence is insufficient to conclude that no substantive opposition existed in the wider historical discussion.

Decision vs debate

The official versions were proposals, not decisions. Their rationales present the Certification Task Force's supporting case and state that there were no opposing arguments, but that wording does not provide a complete historical account of the discussion. The supplied participant record contains three messages from three total participants, two discussion participants and one formal actor; only one stance was classified, and it was procedural or neutral. The only recorded decision was withdrawal on 2011-07-26. Publication of version 4 on 2011-03-21 therefore cannot be treated as acceptance or implementation.

Participation concentration

The measured discussion was sparse and concentrated: activity occurred on two days, the top contributor accounted for two-thirds of messages, the HHI was approximately 0.5556, and the effective participant count was 1.8. One discussion participant contributed a single message and one was a repeat contributor. These figures do not demonstrate broad community consent or opposition. They describe the supplied corpus only, and its completeness is not established.

Chokepoint analysis

The proposed chokepoint was certificate issuance and lifecycle control. Under the later drafts, the RIPE NCC would issue certificates to the registered holder, certificates would reflect the resource's registration status, and eligibility would be limited to members in good standing. Version 2 expressly provided for automatic revocation when the holding record was withdrawn. Version 4 extended the proposed gate to specified IPv4 and IPv6 allocations. This would have made registry status and institutional standing important conditions for certification, but because the proposal was withdrawn, the record establishes a proposed control arrangement rather than an operational one.

Uncertainties

The supplied record does not state why the proposal was withdrawn, who initiated the withdrawal, or whether a successor policy, contract or operational instrument replaced it.
It does not establish whether similar RPKI certification rules were later adopted or implemented through another route.
The completeness of the discussion record is unknown, so the sparse participation metrics cannot be treated as a complete measure of community engagement.
Only one participant stance was classified as procedural or neutral; substantive support and opposition cannot be reconstructed from the supplied stance data.
The operational consequences of certificate revocation for routing, customers or resource use are not supplied.
The definitions and procedures for good standing, deregistration, appeals, grace periods and Certificate Practice Statement discretion are not fully established by the supplied facts.

Source facts

Policy evolution

1.0

Withdrawn · Withdrawn

Unknown

BTW Analysis

What changed

What the debate was really about

Arguments that prevailed

Arguments that did not prevail

Decision vs debate

Participation concentration

Chokepoint analysis

Uncertainties

Source facts

Policy evolution

Debate timeline

1.0

2.0

3.0

[address-policy-wg] Agenda (second draft) for RIPE 61 in Rome

4.0

Re: [address-policy-wg] Legal counsel on 2008-08 (Initial Certification Policy in the RIPE NCC Service Region)

withdrawn

Entities

Sources and coverage

Official policy sources

Public discussion archives

Coverage

Freshness

Known gaps