Summary

  • A 2005 NANOG archive describes a historical late-submission path using Program Committee mailing-list discussion, attempted consensus and a chair decision informed by timeliness and agenda space.
  • Recent official meeting notices provide a more visible urgent route: short Lightning Talks submitted during the meeting under a published deadline.
  • Neither record proves favoritism. Together they show which facts a bounded exception should expose without publishing confidential proposals or reviewer deliberations.

A programme deadline cannot anticipate every incident

Operator meetings deal in events that do not respect conference calendars. A routing leak, software failure, measurement anomaly or emergency coordination lesson may become important after the ordinary call for content closes. If the programme has no late path at all, it can force the most operationally useful material into a corridor conversation.

NANOG's archived January 2005 explanation recognised that problem. It described late proposals as a less formal process: the Program Committee discussed them on its mailing list, tried to reach consensus, and left the chair to decide after considering the discussion, the timeliness of the material and available agenda space. That is a meaningful historical fact. It is not evidence that the same rule remains in force in 2026.

The design contains an intelligible purpose. Relevance can change quickly, and programme capacity is finite. But the same ingredients—urgency, discretion and a scarce final slot—also create a governance question. If the route is invisible, an outsider cannot distinguish a genuine exception from preferential access.

Lightning Talks make part of the late route visible

Official attendee notices for NANOG 90 and NANOG 91 advertised ten-minute Lightning Talks during meeting week. They described a short presentation by an in-person attendee on a topic relevant to the NANOG audience. The notices gave specific submission cut-offs, required an abstract, and said selected speakers would be notified shortly before their session. A NANOG 97 notice continued to direct proposals through the agenda system under the “Lightning Talk” session type.

This is not an invisible exception. It is a publicly announced format with a defined duration and deadline. Nor do the notices show that Lightning Talks bypass review. They show a fast route, not an unreviewed one.

That distinction matters. A standard proposal arriving after the ordinary deadline, a Lightning Talk submitted inside its own published window, and a chair-authorised emergency insertion are three different acts. Treating all three as “late content” hides who was eligible, which capacity pool was used and what minimum decision process applied.

The exception needs a small public envelope

NANOG need not publish a proposal, expose security-sensitive operational details or name individual reviewers. It can define the envelope around the decision.

Before the meeting, the public rule could state the eligible triggers: a materially new operational event, a time-sensitive measurement, or another reason that could not reasonably have met the ordinary deadline. It could identify which format is available, how much unallocated time remains, the minimum number of reviewers, the conflict rule and who makes the final call.

After the meeting, a compact aggregate record could state how many exception requests were received, how many were accepted, which route was used and whether the accepted material displaced a previously scheduled item. Small counts may need suppression, and reasons can be grouped. The purpose is not to create an appeals tribunal. It is to make the boundary of discretion observable.

Capacity should be known before names are considered

Scarcity is easiest to govern when the slot is defined before the applicant. A committee can reserve a small Lightning Talk block, publish its duration and close it at a stated time. If an emergency insertion sits outside that block, the record should say whether it filled unused capacity, replaced a cancellation or displaced another item.

This sequencing reduces the risk that capacity expands for a familiar speaker and contracts for an unfamiliar one. It does not prove that such treatment occurs at NANOG. It is a prospective control for any small technical community in which expertise, reputation and repeated participation naturally overlap.

Conflicts also need a minimum rule. A reviewer closely connected to a submitter or subject can step out, while the remaining reviewers still apply the same relevance and presentation criteria. Aggregate reporting can show that the control operated without revealing the identity of the proposer or reviewer.

Reasons protect discretion rather than abolish it

An exception decision cannot be reduced to a mechanical score. A live operational event may deserve room precisely because its importance was unknowable earlier. The chair and committee therefore need judgment.

The useful record is short: the qualifying trigger, the capacity source, the minimum review completed, any conflict handling, and the bounded reason for acceptance or rejection. These elements preserve judgment while preventing “urgent” from becoming a label that ends scrutiny.

Heng Lu's doctrine is relevant only as an analytical lens here: keep the observation, decision and return path distinct. It is not evidence that NANOG maintains such a ledger today.

A fast lane should be narrow enough to trust

NANOG's public materials do not establish a current secret channel, favoritism or a general right to late acceptance. They document an old discretionary process and recent advertised Lightning Talk windows. The responsible conclusion is therefore institutional, not accusatory.

A late route keeps an operator forum responsive. A published envelope keeps that route from becoming an alternative programme available only to people who know whom to ask. Urgency and equality do not have to be opposites. The strongest design makes urgency admissible, capacity finite, review minimums stable and the final use of discretion visible after the fact.

Sources and limits