Summary

  • NANOG's official 2026 call for content said it sought more than 1,700 minutes across General Session and Breakout Rooms, while official agenda and attendee pages also identify workshops, keynotes, sponsor showcases, lounges and community forums.
  • Scheduled minutes reveal institutional allocation, not actual attendance, quality or influence. A defensible audit must keep those measures separate.

The clock is an institutional choice

Every programme makes choices before anyone enters a room. A thirty-minute plenary prevents another use of those thirty minutes. A multi-hour workshop consumes more clock but may serve fewer people. A parallel breakout expands total supply while forcing attendees to choose. A keynote concentrates attention; a Birds of a Feather session distributes it.

NANOG made the scale unusually visible in a 2026 reminder for NANOG 97. The organisation said it was looking to schedule more than 1,700 minutes of content between General Session and Breakout Rooms and had confirmed 180 minutes at the time of the notice. It also linked requested subject areas to feedback from recent surveys. Those statements establish a large planned content budget and a stated demand signal. They do not show the survey sample, proposal pool or final allocation formula.

Different formats purchase different kinds of value

The official NANOG 97 materials place opening remarks and a keynote on the same public day as technical sessions. Separate notices identify long hands-on workshops, a peering coordination forum, sponsor showcases, a network lounge and social activity. Each format can serve a legitimate purpose, but not the same purpose.

A plenary can create a shared reference point. A breakout can support specialist depth. A workshop trades reach for practice. A BoF can coordinate an emerging problem before it is ready for a formal presentation. Networking may enable the bilateral work through which operators actually resolve interconnection problems. Sponsor visibility may help finance a meeting without proving control of its technical agenda; the reviewed sources do not establish the terms or effect of any particular contribution.

The governance problem begins when these categories are counted as though they were interchangeable. “More than 1,700 minutes of content” is meaningful, but it does not say how many minutes were operational evidence, tutorial instruction, institutional ceremony, commercial access or unstructured coordination. Nor should every networking minute be labelled commercial. The relevant distinction is the stated function and the access rule, not suspicion attached to a room.

Scheduled time is not experienced time

An agenda audit needs at least three clocks. The first is scheduled minutes: what the institution allocated. The second is delivered minutes: what actually occurred after cancellations, overruns and substitutions. The third is participant exposure: it should not be calculated mechanically from scheduled duration alone, but interpreted alongside room capacity, attendance and parallel conflicts.

The public agenda principally supports the first clock. It may sometimes support the second through recordings or revised timestamps. It rarely supports the third without attendance data. A packed fifteen-minute operational talk can affect more people than a long workshop; a small workshop can nonetheless produce more practical value per participant.

Therefore, this article does not rank formats by duration. It argues that duration is a governance input whose meaning depends on reach, purpose and evidence.

A longitudinal ledger would reveal priorities without inventing motives

NANOG's historical archive makes sampling across decades possible. A reproducible ledger could record meeting, date, room, session type, scheduled duration, concurrency, access condition and declared sponsor relationship. Changes in page structure and missing records should remain visible rather than silently filled.

The analysis could then ask whether operational case studies gained or lost plenary time, whether workshops grew, how much technical material moved into parallel rooms, and whether community coordination received stable space. It could also identify ceremonial and explicitly commercial blocks without treating a speaker's employer as proof of commercial content.

This would reveal operating priorities, not hidden intentions. If workshop time expands after repeated attendee demand, that is a testable relationship. If sponsor showcases sit outside technical selection, the separation should be recorded. If the agenda changes after publication, the revision itself is useful evidence.

Editorial craft still needs a reviewable record

The strongest counterargument is correct: programme composition is editorial craft, not an allocation formula. A novel operational incident may deserve a plenary despite a small audience; a recurring tutorial may deserve hours because it builds shared capability. Equal minutes for every topic would be an arbitrary rule.

But judgment becomes stronger when its aggregate effects can be examined. NANOG could publish a post-meeting table of scheduled and delivered minutes by declared format, along with room capacity, cancellations and major substitutions. Attendee survey results could be reported as distributions with response counts rather than as a single mandate.

No proposal-level score or private deliberation is required. The point is to let the community see whether the final portfolio resembles the needs the institution says it heard.

Scarce minutes should have visible jobs

A defensible programme need not maximise technical-talk minutes. It should make the job of each block legible. Technical evidence, instruction, operator coordination, institutional business and sponsored access can all exist, but their boundaries should not depend on insider knowledge.

Heng Lu's doctrine is useful only as an analytical lens: preserve the observed demand, the allocation decision and the later outcome as distinct records. It is not evidence that NANOG currently uses such a ledger.

The conference clock is too valuable to be treated as neutral, and too complex to be reduced to a score. A transparent allocation record occupies the useful middle: it preserves editorial judgment while allowing the community to see what that judgment repeatedly buys.

Sources and limits