Summary

  • npNOG’s public websites preserve a useful record of programmes, teaching materials and event continuity, but they also contain date conflicts, unfinished download promises and pages whose headings outlive their visible data.
  • Those gaps do not prove that an event failed, a presentation never existed or an attendee list was deliberately removed. They show why a technical community needs named archive custody, a correction log and a durable export of its public record.

The archive already does real institutional work

A network operators’ group leaves more than a sequence of meetings. Its programmes show what engineers considered urgent. Its workshop pages preserve the order in which skills were taught. Speaker lists, fellowship notices and committee pages show who was visible, who received support and how work was divided. Downloadable slides can outlast the room in which they were presented and become teaching material for people who never attended.

npNOG has preserved enough of that record to make serious inquiry possible. Its main Events page links multiple numbered editions. The npNOG1 site still exposes named files from a small-office and home-office network-design workshop, including material on addressing, cabling and layered network concepts. It also links four conference presentations, covering international connectivity, IPv6, information security and a DNS denial-of-service experience.

That survival deserves credit. Volunteer organisations often move between organisers, hosting arrangements and site templates. A decade of accessible pages is not automatic. The right starting point is therefore not that npNOG has no archive. It is that the archive has become important enough to deserve explicit stewardship.

A date conflict is a correction problem, not an erased event

The official Events index contains internal contradictions. Its npNOG5 entry says the event ran from 8 to 14 December 2018, then describes the conference as taking place on 13 December 2019. Its npNOG2 entry says the event was held from 15 to 17 June 2016, then places the workshops and conference in June 2017.

These inconsistencies are visible on npNOG’s own page. They should be reported as inconsistencies, not silently harmonised by a writer choosing the date that looks most plausible. At the same time, they do not establish that either event failed to occur. Other official pages, programmes, photographs or external records may identify the correct dates. The archive problem is narrower: a reader should not have to reconstruct which part of an official sentence controls.

A correction log would solve more than the two examples. It could state the original text, the corrected value, the date of correction and the record used to support it. The old wording need not disappear. Preserving the change would let future researchers distinguish an error in the event from an error in the archive.

That distinction matters for institutional accountability. Editing a page is a technical act. Deciding that one record is authoritative is a governance act. The public site does not identify who performs that second role, what evidence they require or where a disputed correction can be reported.

“Coming soon” can become part of the permanent record

The npNOG1 Downloads page demonstrates both preservation and incompleteness on the same screen. It provides working categories and named files for one workshop and several conference sessions. The Optical Fiber Network Design Workshop, however, still reads “Coming soon”. The conference section says that more material will be added soon.

The npNOG2 Downloads page is thinner. It says that presentations, slides and other programme material will be available and that the page will be updated during and after the programme. In the inspected public page, no itemised download inventory appears beneath that promise.

Neither page proves that the absent files were never created. A trainer may have declined publication. A licence may have been unclear. A file may have lived on a drive, a mailing list or a platform that was not migrated. The workshop itself may have been fully delivered. “Not visible in this public page” is the supportable finding; “did not exist” is not.

Yet an open-ended promise creates a real archival ambiguity. Is the page unfinished, or is the absence intentional? Was permission refused, or was follow-up lost? Is a file available somewhere else? A short status beside every expected item would answer those questions without demanding that npNOG publish material it has no right to distribute. Useful labels could include published, not supplied, publication declined, link lost, superseded and status unknown.

An empty list is not an empty room

The npNOG11 Attendees page renders an “Attendee List” heading but no visible attendee entries in the inspected page. That is an observable condition of the page, not evidence about attendance.

There are several plausible explanations. The list may never have been populated. It may have been removed for privacy. A database query or plugin may no longer return its rows. The content may have been intentionally withheld. Without a note from the archive owner, the public cannot distinguish among them.

Privacy also changes what good preservation means. A complete public archive is not necessarily one that retains every participant’s name forever. Registration records can contain personal data, employment information and travel details that should not be exposed merely because an event happened years ago. The better archival question is whether the public page accurately describes what it preserves.

If names were removed, a note could say that the list is no longer public while retaining a privacy-safe aggregate. If a technical failure hides the entries, the page could report the failure and the last known good capture. If no public list was ever intended, the heading should not imply otherwise. Accuracy can improve without republishing personal data.

The site does not say who owns the memory

The inspected pages identify npNOG and its individual editions, but they do not publish an archive-custody rule. A reader cannot see who is responsible for preserving files, resolving contradictory dates, maintaining domains, exporting content before a platform change or deciding how long participant information remains visible.

That absence should not be converted into a claim that nobody performs the work. Volunteers, committee members or a hosting provider may already maintain backups and fix reported errors. The public evidence simply does not assign the responsibility.

Named custody is important because a multi-site archive has several failure points. Domains expire. Themes and plugins stop working. External slide links disappear. A committee changes and the person holding the only copy leaves. A later organiser may inherit a website without the credentials, consent records or file licences needed to repair it.

The durable control is not a particular content-management system. It is a handover package: current domain and hosting responsibility, an export of pages and media, checksums for preserved files, licence or consent notes, a list of known gaps, and a route for reporting corrections. Copies can then move between platforms without making one vendor, volunteer or password the sole custodian of the historical record.

A minimum archive standard can remain proportionate

npNOG does not need a museum-scale preservation programme. A lightweight standard could cover the public evidence most likely to matter later.

First, each edition could have a manifest listing its dates, location, programme, committees, workshops, speakers, fellows where publication is appropriate, and available presentation material. Every field could accept “unknown” or “not public”. Completeness would mean that the status is explicit, not that every field is filled.

Second, the organisation could publish a correction log. A correction should identify the affected page, old value, new value, supporting record and date. Disputed facts—such as edition labels used differently by partner institutions—could remain disputed rather than being flattened into one retrospective answer.

Third, public files could be exported periodically with stable names and checksums. A checksum does not prove that a slide is accurate, but it lets a later custodian verify that the archived file has not silently changed.

Fourth, privacy and publication rights should be recorded alongside the archive. Attendee identities, photographs and presentation decks do not all have the same permission basis. A retention decision should follow that basis, not an assumption that “public once” means “public forever”.

Finally, the public should have one correction route and one accountable role, even if the person holding that role changes. The role could acknowledge reports, preserve the evidence considered and state whether a correction was accepted, rejected or left unresolved.

Imperfect preservation is still preservation

The strongest counterargument is practical. npNOG is a technical community, not an archive institution. Volunteers should spend scarce time teaching, convening and solving operational problems, not rebuilding old websites. Some missing files may have no surviving licence or copy. Publishing detailed participant records may create more risk than value.

That objection is persuasive against perfectionism. It is not an argument for ambiguity. A one-page manifest, periodic export and correction log are modest controls. They reduce repeated research, prevent later organisers from inheriting unexplained gaps and let the community decide deliberately what should remain private.

The surviving record already demonstrates why the work is worthwhile. The named npNOG1 files give later readers evidence of what one workshop taught. The Events page provides a chronology even where individual dates need correction. The unfinished and empty pages reveal not that the meetings lacked value, but that institutional memory requires maintenance after the event team has moved on.

npNOG has preserved enough of its first decade to make its history visible. The next step is to make the preservation process visible too: who keeps the record, how errors are corrected, what is intentionally withheld and which gaps remain unknown.

Sources