Summary

  • RFC 9775 applies common expectations across IRTF meetings, research-group and IRTF-wide forums, yet it assigns moderation, approval, confidential reporting and appeal to different roles.
  • A warning, temporary posting suspension, permanent restriction, membership decision, ethics review and complaint do not share one decision chain simply because they appear in the same code.
  • A local forum-jurisdiction receipt can record the applicable venue, policy, authority, approval, scope, expiry and appeal without copying a confidential allegation. This is Daniel Kade’s editorial proposal, not an RFC field or case registry.

Imagine that an identical disruptive message lands on two mailing lists. One belongs to a research group; the other is an IRTF-wide forum. A reader sees the same words. The code supplies the same expectation of dignity and professional conduct. Yet the authorized moderator is different: the research-group chairs moderate their group forums, while the IRTF Chair moderates the IRTF-wide forum. Governance begins at that difference.

The temptation is to call the code “the authority.” A text cannot warn a participant, suspend posting rights, keep a report confidential or hear an appeal. People in specified roles do those things, under a policy that applies to a particular activity. If the venue is misidentified, an otherwise reasonable decision may be made by the wrong role. If the decision is not bounded by time and scope, a temporary intervention can harden into an unreviewed permanent sanction. If an appeal returns to the same conflicted decision-maker, the nominal remedy is not operational.

RFC 9775 makes these distinctions unusually visible. Published in March 2025 as an Informational document in the IRTF Stream and produced through the Internet Research Steering Group, it is not an IETF product or Internet Standard. The IRTF and IETF cooperate and often share infrastructure or meeting space, but the IRTF exists for longer-term and sometimes speculative research rather than the IETF’s standards consensus and near-term deployability. The code therefore belongs to an identifiable institution and process, not to an abstract “Internet community.”

One expectation, several operating surfaces

The code reaches research-group meetings, open meetings, conferences, workshops, mailing lists, virtual meetings and other online forums. It asks participants to treat one another with dignity, respect and courtesy. It also addresses research integrity, research ethics, inclusive language, accessibility and participation. Those subjects sit in one document, but they are not one administrative object.

On a group list, the research-group chairs have the moderation role. On an IRTF-wide list or forum, that role belongs to the IRTF Chair. A moderator may warn someone or temporarily suspend posting rights. A permanent posting suspension requires approval from the IRTF Chair. The escalation changes the authority basis: duration is not merely a larger value in the same moderator’s form; permanence introduces another decision point.

That separation protects both participants and moderators. A chair can contain an immediate disruption without pretending that a rapid intervention has already satisfied the standard for an indefinite restriction. The affected participant can ask who approved the continuing measure. Other participants can know that the forum is being moderated without learning confidential details that do not belong in a public log.

The same distinction appears in appeal. A moderation decision can be appealed to the IRTF Chair. If the IRTF Chair made the challenged decision, the concern goes to the Internet Architecture Board. The handoff matters because review must remain capable of contradicting the original authority. “Appeal available” is not an adequate record when the record cannot say where the appeal went or whether a conflict changed the destination.

Confidential reporting is a route, not a data-export instruction

Harassment concerns may be raised with the relevant research-group chair, the IRTF Chair or the Ombudsteam. The Ombudsteam operates independently and confidentially. That choice of route is part of the protection: a person may need a channel outside the ordinary forum hierarchy, especially when the problem concerns someone inside it.

Confidentiality does not mean that governance becomes unverifiable. It means verification must be deliberately narrow. An operational system may need to know that a qualified route received a concern, that an authorized role imposed a bounded action, that an approval was obtained and that an appeal destination exists. It usually does not need the complaint narrative, private evidence, protected characteristics, witness names or a copy of confidential correspondence.

This is why a generic ticketing system is dangerous. It rewards completeness measured by fields filled and attachments retained. The result can be a durable dossier available to many operators, even when the conduct process promised a protected channel. Accountability asks for proof of the authority chain and state change. Surveillance asks for everything. The two should not be confused.

The IETF’s anti-harassment policy statement and the Ombudsteam contact page help explain the surrounding reporting architecture. They do not turn every recipient into the same adjudicator, and they do not eliminate the need to identify which policy applies to an IRTF forum.

Co-location makes policy notice a governance act

A workshop held alongside another conference can look like one event while remaining subject to a different host policy. RFC 9775 allows a co-located activity to follow the host event’s anti-harassment policy, provided organizers confirm its acceptability with the IRTF Chair and make the applicable policy clear to participants.

Notice is not decoration here. It is the first execution of the jurisdiction choice. A participant cannot make an informed decision about attendance, reporting or appeal if the applicable rule becomes visible only after an incident. Organizers likewise cannot combine the strongest clause from one policy with the most convenient authority from another. The policy version, host, approval and participant notice must travel together.

Shared systems create another boundary. Exceptional legal advice may require restricting an individual’s access to IETF and IRTF IT systems or activities. The IRTF Chair acts in consultation with the IESG because an access restriction on joint infrastructure or meetings may affect the standards process. The IESG statement on restricting access describes that exceptional context. It should not be read as a general shortcut around ordinary moderation.

Conduct is not research integrity, and neither replaces ethics review

RFC 9775’s research-integrity provisions concern honesty, attribution, authorship and the accurate presentation of evidence and consensus. Plagiarism, falsification or misrepresentation can justify retraction or exclusion. A generative AI system must not be listed as an author. Significant generated content may be used but must be disclosed; spelling, grammar, translation and presentation assistance need not be disclosed under the document’s stated rule.

These are provenance and integrity questions, not merely tone questions. A courteous paper can still fabricate evidence. A contentious discussion can still contain valid research. Moderating a list post does not resolve whether a publication should be retracted, and an authorship declaration does not decide whether a participant may attend a meeting.

Research involving humans or personal data introduces a third chain. Researchers should consider risks and benefits, respect persons, applicable law and the public interest, and consult an institutional review board, research ethics committee or equivalent body when ethical concerns may arise. The IRTF code does not replace those bodies. Their institutional competence cannot be recreated by asking a mailing-list moderator to approve the study.

The older IRTF process documents help locate these boundaries. RFC 2014 established the IRTF’s research orientation. RFC 7154 states conduct expectations, while RFC 7418 discusses IRTF research-group guidelines and procedures. RFC 7776 addresses anti-harassment procedures, and RFC 7841 records the IRTF Chair’s model. The point is not to collapse those sources into one super-policy. It is to see why roles, purposes and remedies need explicit joins.

Membership and participation have their own reversibility

Research groups can be open or have limited membership. When membership is limited, the limitation must be fair and transparent and have a rationale. Concerns and appeals go to the IRTF Chair. Membership is therefore neither a favor wholly beyond review nor the same thing as posting permission on one list.

Accessibility makes the difference practical. Participation includes clear communication for people who are not native English speakers, accessible presentation materials, usable remote participation and support mechanisms. The IRTF’s inclusive-language guidance is part of that operating environment. A formally open meeting that cannot be heard remotely, navigated with assistive technology or understood without unexplained idiom is not made inclusive by the word “open.”

RFC 8890 argues that the Internet is for end users. In this setting, that principle directs attention from policy publication to the participant’s actual route: could the person discover the applicable rule, reach the right channel, receive notice of an action and use a meaningful appeal? Institutional legitimacy is experienced at those interfaces.

A forum-jurisdiction receipt records the handoff, not the allegation

The proposed forum-jurisdiction receipt is a small local record for this chain. It begins with the activity and venue: research-group meeting, group list, IRTF-wide forum, co-located event, membership decision, publication process or shared-system restriction. It identifies the policy and version participants were told would apply, plus the evidence of notice where the venue choice required it.

The receipt then records the acting role and authority basis, not merely a person’s name. It classifies the action as interim or final and captures its scope, start, duration and expiry. A permanent posting restriction records the IRTF Chair’s required approval. A confidential concern records the protected route and a non-sensitive reference, not its narrative. Notification and appeal destination are explicit. If the original decision-maker has a conflict, the substitution is visible. Outcome and expiry close the state transition.

Such a receipt must be local, access controlled and short lived. It must not become a public complaint list, an IRTF-wide case database or a second archive of Ombudsteam material. It should refuse unnecessary identities and private evidence. Its usefulness lies in answering bounded questions: which rule governed this forum; who could take this action; which additional approval was required; where could it be reviewed; and did the temporary state actually end?

This is an editorial method, not a field specified by RFC 9775. The RFC Editor record classifies the document and the published RFC text supplies the normative source for its statements; neither mandates this receipt.

A published code is specification, not observed outcome

RFC 9775 has no IANA actions and does not directly change Internet security. It asks researchers to consider downstream security and privacy consequences. More importantly for governance, it does not prove that a participant saw the relevant policy, a moderator used the right authority, confidentiality held, a temporary measure expired or an appeal reached an independent reviewer.

Heng Lu’s distinction among specification, localized future decision, voluntary adoption, actual execution and observed outcome prevents that category error. “Minimum Initial Specification, Localized Future Decision, Voluntary Adoption” explains why a common text can enable coordination without predetermining every future local act. “On Why BTW Media Exists” separates the event worth observing from the advocacy surrounding it. “The Policy Mirror” warns against treating policy language as proof of operational reality.

The right conclusion is not that authority should be centralized so the map becomes simpler. The IRTF’s forums, research judgments, confidential channels, ethics institutions and shared infrastructure have different purposes. The discipline is to preserve those differences while making every handoff legible. A shared rule earns legitimacy when each participant can know who may act here, on what basis, for how long and before which reviewer.

Sources