Summary

  • Nominations for the 2027–2029 IRTF Chair term close at 23:59 UTC on 8 September 2026. Community input runs to 5 October, and the IAB plans to announce its selection on 21 October.
  • The IAB estimates that the role consumes about 25% of normal work weeks and becomes full time for about six weeks around the ANRW, ANRP review, IETF meetings and IAB retreats.
  • RFC 7827 allows some administrative duties to be delegated, but says responsibility ultimately remains with the Chair. The office therefore needs more than a single assurance of general availability.
  • Daniel Kade proposes a public two-speed capacity map: baseline duties, surge windows, movable work, delegation, retained accountability, collision rules and backup. It can protect private candidate material while making the design of the office testable.

A deadline with two clocks

The Internet Architecture Board's call is precise about the calendar. Nominations, including self-nominations, close at 23:59 UTC on 8 September. Candidate names are to be published together when the nomination period closes. Community input continues until 5 October, the IAB intends to announce the incoming Chair on 21 October, and the two-year term begins at IETF 128 on 8 March 2027.

It is also precise about time in a more useful sense. The job averages about 25% during normal work weeks, the call says, but rises to full time for about six weeks a year. It names the pressure points: the Applied Networking Research Workshop, the review period for the Applied Networking Research Prize, the three IETF meetings and IAB retreats.

Those are not two descriptions of the same burden. They are two operating regimes. A person can have one day a week available across a quarter and still be unable to clear a continuous week for an IETF meeting. Someone else can protect the travel and event weeks but lack the recurring attention needed for documents, Research Group questions and monthly reporting. An annual average makes those two failure modes look identical.

The nomination form compresses the distinction. Candidates are asked to confirm their willingness and ability to devote the level of time required. They also provide a biography and a statement of interests, concerns, goals and motivation. With the exception of names, the information supplied to the IAB remains confidential. That privacy rule is legitimate; it prevents the selection from becoming a public release of personal dossiers or third-party comments. It does not require the operating assumptions of the office to remain implicit.

This is not a ceremonial chair

The load matters because the office carries decisions, not merely invitations. RFC 2850 gives the IAB authority to appoint and remove the IRTF Chair for a renewable two-year term. The Chair manages and organizes the Internet Research Task Force and joins the IAB ex officio.

RFC 7827 divides the work into strategic and administrative domains. Strategically, the Chair helps determine which Research Groups should be chartered, which have ceased to be viable, and which research communities should be brought into contact with Internet engineering and operations. A decision to encourage a new group or close an exhausted one shapes which questions receive an institutional home, access to meetings and a publication route.

The administrative list is longer than the title suggests. The Chair follows IRTF Stream documents, ensures review, coordinates required IESG review and follows up with IANA and the RFC Editor. The office manages Research Group chairs and the IRSG, handles meeting slots and materials, keeps web and Datatracker state aligned, maintains operating procedures, reports monthly to the IAB and organizes the IRTF Open Meeting.

The maintained IAB task inventory widens the picture further. It estimates 20–25% FTE plus at least four long-haul trips a year and the time needed to keep abreast of research communities. Its list runs across the publication stream, existing and proposed Research Groups, ANRW, ANRP, IAB participation, communications, diversity work, process upkeep and conduct matters.

The wiki is not an RFC, and a task list is not a time sheet. But it demonstrates why a single percentage is a poor control surface. Recurring work, scheduled surges and discretionary interventions consume attention differently. A stalled document can wait a few days. A meeting cannot. A sensitive Research Group closure may expand without warning. An IAB retreat and an urgent stream decision can occupy the same calendar even though neither was misestimated in isolation.

Delegation moves work, not responsibility

RFC 7827 supplies the crucial governance boundary. Some administrative duties may be delegated permanently or temporarily to other IRSG members, but responsibility ultimately remains with the IRTF Chair.

That sentence prevents two opposite errors. The first is heroic-office design: pretending that a chair must personally execute every review, agenda action and coordination task. The second is accountability evaporation: treating delegation as proof that the office no longer needs capacity for supervision, integration or correction.

A delegated review still needs an owner, a deadline and an escalation path. An assigned event task can reduce execution time while increasing coordination time. When a delegate becomes unavailable, the obligation returns to the office whose name remains on the process. Capacity therefore has at least three components: work the Chair must perform; work others may execute while the Chair remains accountable; and surge work that displaces everything else.

The distinction is especially important for strategic decisions. Research Group chartering and closure, appointment of chairs, interpretation of process questions and representation on the IAB are not interchangeable units of clerical effort. They require judgment, relationships and attention at the time the decision arises. A nominally available quarter-time schedule cannot be banked and redeemed after the decision window has closed.

Publish the capacity design, not the private dossier

The IAB can make the appointment more testable without exposing confidential material. It should publish a two-speed capacity map for the office and ask the selected Chair to acknowledge that map in a short, candidate-approved statement.

The baseline half should identify recurring duties and the assumptions behind the 20–25% estimate. Which work is weekly or monthly? Which queues can move? Which tasks depend on IETF meeting cycles or publication traffic? The surge half should name known full-time windows, their likely duration and the obligations that cannot safely be deferred.

A delegation column should then distinguish execution from accountability. It should name the class of person or body that may take a task, not disclose private staffing arrangements. A retained-owner field should show where responsibility returns. Collision rules should say what happens when a Research Group decision, document deadline and IAB obligation overlap. A backup field should identify the escalation route for illness, employer conflict or an unexpected event. Finally, the map needs a review date, because a workload estimate that was accurate for one portfolio can silently expire as groups and publication queues change.

The selected Chair's public acknowledgement can be narrow: access to both baseline and surge capacity has been considered; any planned delegation fits the published boundary; and material inability to sustain the role will be raised with the IAB. It need not name an employer, reveal leave arrangements, reproduce nomination answers or expose community feedback.

This proposal does not imply that the incumbent or any prospective candidate lacks time. Dirk Kutscher has indicated interest in another term, but the checked record does not establish the final candidate slate or an outcome. Nor does it prove that the IAB currently ignores peak capacity. The problem is evidentiary: outsiders can see the two demands in the call, but not the durable operating record that keeps them separate after appointment.

Sources