Summary

  • npNOG describes itself as an open forum and says anyone may join and participate by complying with its terms.
  • Its mailing list is described as open and “sometimes moderated”.
  • Seven mailing-list rules define relevant and prohibited content, but the reviewed terms do not name the moderator, activation trigger, notice practice, appeal route or recusal rule.
  • The Code of Conduct reaches online and digital spaces, yet its conference-sanction clause does not establish a list-specific moderation procedure.
  • Publishing the moderator role, intervention triggers, notice and review route would make enforcement legible without weakening anti-spam or anti-harassment controls.

Openness is a rule about access

npNOG presents itself as an open forum for the technical and operational Internet community in Nepal. Its About page says anyone is free to join and participate if they comply with the organisation's terms and conditions.

The page makes the same promise more specifically about the mailing list. It calls the list open, says anyone may join, and describes it as a place to ask questions, distribute information and share operational knowledge.

That is more than promotional language. A network-operators list can be part of the working infrastructure of a technical community. It can help an engineer test whether an incident is local, circulate a change that affects interconnection, find peers who have seen the same failure, or introduce a practice that does not yet have a formal venue.

Access to that channel therefore has governance value. The list does not need to accept every message. It does need a visible relationship between its claim of openness and the authority that limits distribution.

“Sometimes moderated” adds a gate

The same About page qualifies openness with two words: sometimes moderated.

Moderation can be entirely legitimate. An operational list can be damaged by product marketing, personal attacks, automated replies, repetitive noise or material far outside its purpose. A community that promises an open forum does not promise an unfiltered one.

But moderation introduces a gate between a participant's message and the community that may read it. Depending on the design, that gate might hold a first post, screen every post during an incident, reject prohibited material, remove content after publication or operate mainly as an anti-spam filter.

The reviewed pages do not say which of these models npNOG uses. They do not identify the condition represented by “sometimes”, or whether the condition is attached to a person, a message, a period of heightened risk or an automated system.

That absence is not evidence of secret censorship. It is a boundary on what a participant can learn from the published rule.

The content boundaries are more visible than the decision path

npNOG publishes a short mailing-list terms page. It says discussion should focus on Internet operational and technical issues. It prohibits product marketing, disrespect, foul language, character assassination, political, philosophical and legal posts, private marketing use of the list, and auto-responder messages.

Those rules supply a real substantive boundary. A poster can see that the list is not a general-purpose political forum or a sales database. Participants can also expect protection from personal abuse and automated noise.

The page is less specific about enforcement. In the reviewed text, it does not name a moderator or committee. It does not say whether moderation occurs before publication or after a complaint. It does not describe what happens when a message crosses a line, whether the poster receives the relevant rule and a reason, or whether another person can review a disputed decision.

Some of the listed boundaries also require judgement. Product marketing can be obvious, but an operator describing a vendor-specific failure may need to name a product. A legal development can have direct routing, licensing or cybersecurity consequences even though “legal” posts are prohibited. Respect is essential, but reasonable criticism can be uncomfortable without becoming personal abuse.

Rules that need interpretation make the decision path more important, not less.

The Code of Conduct does not fill the list-specific gap

npNOG also publishes a Code of Conduct and anti-harassment policy. It says conduct expectations apply at npNOG venues, related social events and all online or digital spaces. It directs a person with concerns to any member of npNOG staff.

That broad reach is useful. It prevents online participation from being treated as outside the community's safety expectations.

The policy also says conference participants who violate its rules may be sanctioned or expelled without a refund at the discretion of conference organisers. That identifies an authority and consequence in an event setting.

It should not automatically be projected onto the mailing list. “All online/digital spaces” establishes scope, but the conference-sanction clause does not tell a list subscriber who approves a held message, whether an email is rejected under mailing-list terms or the harassment policy, or how a list-specific intervention can be reviewed.

The public documents may overlap. The reviewed pages do not publish the bridge between them.

An unnamed role weakens both accountability and protection

Naming a moderation role is not the same as publishing a person's private contact details. A policy can say “mailing-list moderators”, “the list operations team” or “two members designated by the Core Committee”. That is enough to show where authority sits.

Without that role, a participant cannot distinguish an institutional decision from a delivery failure, an automated filter or an individual administrator's action. The community also loses a clear place to send a report that belongs specifically to the list.

The ambiguity can burden moderators too. If the public rule does not explain their mandate, any intervention can appear personal even when it follows a consistent internal practice. A visible role and procedure protect the people doing difficult moderation as well as those subject to it.

The evidence does not show that npNOG moderators have acted inconsistently or that any participant has been harmed. It shows that the public architecture does not let an outsider test consistency.

Notice is the smallest useful record

Moderation need not become a courtroom. A useful notice can be short: the message was held or rejected; the applicable rule was identified; and the poster was told whether a revised post would be accepted.

For spam, phishing or automated abuse, individual notice may be unnecessary or unsafe. A policy can state that exception. For a good-faith technical post, a reason helps the author adjust and protects the forum from repeated boundary disputes.

A second review can also be narrow. It need not reconsider every editorial choice. It can be available where a participant says the moderator applied the wrong rule, misunderstood a technical context or had a conflict of interest.

The source pages do not establish that npNOG lacks notice or review internally. The point is that the public terms do not tell participants whether those safeguards exist.

Aggregate reporting can preserve privacy

The community does not need a public archive of rejected messages. That could repeat harassment, expose security material or turn moderation into spectacle.

An aggregate ledger is enough to make the gate more legible. npNOG could report the number of messages held, released after review, rejected as marketing or automation, rejected for conduct, and reversed after a second look. A zero is useful data too.

The report could exclude message text, reporter identity and private deliberations. It would answer a narrower governance question: how often did the “sometimes moderated” power operate, and under which published boundaries?

If automated filtering accounts for most interventions, that fact should be separated from human editorial decisions. The two mechanisms have different failure modes and should not be hidden inside one number.

A five-part moderation note would close most of the gap

npNOG already publishes the hardest foundation: the purpose of the forum and the substantive rules of participation. It does not need a large constitution for a mailing list.

A short moderation note could add five elements. First, name the role authorised to moderate. Second, state when pre-publication or incident-specific moderation is activated. Third, explain the usual notice and reason practice. Fourth, provide a second-review and conflict-handling route. Fifth, commit to privacy-preserving aggregate reporting.

None of those controls requires npNOG to distribute marketing, tolerate harassment or reveal sensitive reports. They connect enforcement to the openness claim.

An open forum is not defined by the absence of boundaries. It is defined by whether people can see the conditions of entry, the authority that applies them and the route available when that authority may have made a mistake.

npNOG's public pages already show the conditions. Naming the moderation chain would complete the rule.