Summary

  • NANOG's founding record and current list description define an operational exchange surface, not a policy-making chamber for independent networks.
  • Public archives make technical claims and corrections inspectable, but posting volume, silence and thread closure do not establish representation or assent.

The list's strongest claim to legitimacy is functional. In an archived 1993 proposal for the group that became NANOG, the stated work included exchanging technical information and coordinating implementation issues among network service providers. The same proposal put bilateral or multilateral agreements about policies and service features outside the group's objectives. The boundary was unusually clear: cooperate where an interconnected system demands cooperation, without pretending that the forum governs each participant's network.

The current list service preserves that operating character. It names routing, outages, performance measurement, exchange points, traffic engineering, security and broad engineering problems among appropriate subjects. These are areas in which one operator's evidence can save another operator hours, or where several networks may need compatible action to restore service.

The archive shows the distinction in practice. After a major disruption in 2001, a thread collected concrete observations about outages, emergency plans, NOC-to-NOC communication, mutual aid and preventive changes. In a later RPKI discussion, operators compared repository and validator evidence, reported a symptom and preserved subsequent technical clarification. Neither thread needed a plebiscite to be useful. Its value came from testable facts, identified systems and corrections that others could inspect.

Scale does not change that into representation. A 2008 administrator note referred to more than 10,000 subscribers while warning that scores of posts could still concern only a small part of the audience. A 2022 reminder addressed more than 12,000 engineers and again tied the list to technical exchange. Those numbers are dated, attributed estimates, not a current membership audit. More importantly, distribution is not participation. A delivered message does not show that it was read, understood, accepted or authorized for implementation.

Silence is even less specific. It can mean agreement, irrelevance, overload, caution, absence or a decision to solve the problem elsewhere. A long thread may reflect many independent operators, or a handful of persistent correspondents. An archive counter measures visible activity under its own rules; it does not identify all affected networks or the employers that authorized action.

That does not make consensus language forbidden. It makes the referent important. “Several posters converged on this diagnosis” is an auditable claim. “The NANOG community decided” requires a defined decision process and constituency. “Operators adopted the fix” requires implementation evidence. These sentences describe different objects and should not borrow one another's authority.

The mailing list is most credible when it remains an operations room: evidence enters, rival explanations are tested, corrections remain visible and operators decide what applies to systems they control. Its public record can discipline claims without becoming legislation.

Sources