Summary

  • NANOG's official records separately count attendees, survey respondents, members, eligible voters and votes cast. Those populations should not be collapsed.
  • Attendance demonstrates participation and access to expertise. It does not, without further evidence, confer authority from the region's networks or absent operators.

Five populations, five different claims

An official NANOG announcement says 806 attendees gathered for NANOG 96. Another NANOG notice reported 147 survey responses across the three days of NANOG 97 at the time it was sent. The 2020 election page records 553 eligible voters and 176 votes cast. These figures come from different events and years. They cannot be divided into one another to manufacture a participation rate.

Their proper use is more modest and more informative. Attendance counts presence at an event. A survey response counts a response to a particular instrument. Eligible voters belong to a defined governance constituency. Votes cast record action inside that constituency. The 2016 election notice makes one boundary explicit: only NANOG members were eligible to vote and serve on the Board and committees.

NANOG 97 notices also described virtual participation through streaming, chat and submitted questions. Remote viewers are participants too, but their access mode does not by itself establish membership, voting rights or an operational mandate.

A forum is valuable without being a parliament

NANOG's own language describes a forum for North America's operator community. That is a substantial institutional role. A forum can spread operational knowledge, surface incidents, build professional trust and make coordination possible across competitors. None of those benefits requires the meeting to impersonate a regional electorate.

Nor does NANOG claim sovereignty over North American networks. The point is not to rebut a claim the organisation did not make. It is to keep downstream language disciplined. “The community decided” is stronger than “members voted.” “Operators support” is stronger than “attendees responded.” “North America wants” is stronger still.

Speaker lists need the same care. A speaker's employer can identify where that person works; it is not evidence that the employer, its customers or its sector authorized the statement. Operators who did not attend remain part of the regional reality even when they leave no trace in the meeting record.

A participation ledger would preserve the distinctions

A compact post-meeting ledger could report, with dates and definitions, registered attendees, in-person check-ins, virtual accounts or viewers, survey invitations and responses, current members, eligible voters, ballots cast, speakers and represented employers. Unknown fields should remain unknown. Historical counts should keep their original event and year.

Such a ledger would not diminish informal coordination. It would stop one visible population from silently standing in for another. Claims could then name their basis: a member vote, an attendee poll, a programme-selection process, an open comment period or an organizer decision.

Heng Lu's doctrine is useful here only as an analytical lens: preserve observation, constituency, decision and outcome as separate records. It is not evidence about NANOG's present practices.

The meeting room matters precisely because it is not the continent. Its value comes from concentrating expertise and enabling work, while honest denominators show exactly whose participation each record represents.

Sources and limits