Neutral Point of View

Neutral point of view (NPOV) means representing a subject fairly, proportionately, and without editorial bias.^[1] Facts are stated as facts; opinions are attributed to whoever holds them.

Ecosystem and project pages

NPOV matters most where a page describes a project whose own team may be editing it – a structural conflict of interest, covered in full by conflict of interest. Such pages must avoid promotional language ("revolutionary", "the best"), attribute claims about traction or performance to sources, and carry a Written like an advertisement notice banner when they drift into marketing.

Neutrality also applies to what a page leaves out. A project page that records only launches and grants, and never a dormant product or a closed one, is not neutral – the status field and the prose should agree, and both should follow the evidence.

Status is a measurement

Every Ecosystem page carries a status field, and it is where neutrality is easiest to lose. A directory that records launches and grants but never a dormant product or a closed one describes an ecosystem in which nothing ever stops – an omission rather than a falsehood, and still not neutral. A status is therefore set from dated evidence a reader can repeat, and the prose is expected to agree with it.

Radix Rolodex was moved from active to dormant on three readings rather than on a view about the project: its newest published card and deck are both dated 14 February 2026, its deployed bundle was last built on 10 February 2026, and the service is still up with every embed on this wiki still rendering. Nothing about it is broken. It has stopped changing, and the page says only that.

The same discipline stops the field travelling too far the other way. Hermes Protocol reads dormant on four dated checks, and the page states in the same breath that no wind-down was ever announced – closed would be a stronger claim than the evidence carries. Where the evidence is thin, the neutral move is to name the check and its date, not to round the status to whichever end of the scale reads more decisively.

Where a page has drifted into its subject’s own framing, add the Written like an advertisement banner rather than quietly rewriting it, and say in the revision message what was promotional about it – the banner is a request for review addressed to the next editor, and the revision history is public.

When accounts conflict

Status fields go wrong slowly. Contested events go wrong fast, and they are where neutrality is hardest to hold, because the pressure is not to flatter a subject but to settle a question the sources have not settled. The rule is the same one in a harder case: state as fact only what a reader can check for themselves, attribute everything else to whoever said it, and do not resolve a disagreement on the page that is unresolved off it.

Three practices follow from that, and the August 2026 asset drain and network halt is the worked example for each, because this wiki wrote it up while it was still happening. The network was down for 254 hours and restarted on 11 September 2026.

Officialness is not evidence. A statement from the organisation closest to an event is a source, not an adjudication, and it carries no more weight than a source a reader can verify. On 2 September 2026 the Radix Foundation's own channel said that it and the Accountability Council have halted the Radix network. Three hours earlier, in a different channel, Astrolescent's Timan wrote that the node runners made that decision to protect the network. That was not a centralized decision. The ledger agrees with the second: rounds stopped when stake fell below the threshold, which is a thing node runners do individually and no organisation can do to them. The page records both statements, dated and attributed, and says which one the ledger supports. It does not delete the first for being wrong, and it does not adopt it for being official.

Carry the hedge. Where a source qualifies a claim, the qualification travels with it. The first public technical accounts of that incident's root cause came from a protocol developer who twice asked to be taken with a pinch of salt and expected a more accurate write-up later. Repeating such a claim without its hedge converts a working hypothesis into a finding, which is a failure of neutrality as much as of accuracy, and it is a failure this wiki commits on the source's behalf rather than the source's own.

Separate what is checkable from what is asserted. The same page states the mechanism of the flaw flatly, because it is read from published source at a named version and any reader can open the file; it attributes the accounts of that mechanism given in chat, because those rest on the standing of the people who gave them. Both belong on the page. Only the first is stated as fact. Where an event has no checkable layer at all, every account of it is attributed, and the page says plainly that nothing independent settles it.

Neutrality here is not even-handedness for its own sake. Two accounts of an event are rarely equally supported, and saying so is part of the job; what is not part of the job is choosing between them on the strength of who is speaking, or quietly leaving out the version that is inconvenient for either party. See also verifiability, which governs what counts as checkable, and no original research, which governs how far a page may reason from primary sources on its own.

See also

• Verifiability
• No original research
• Notability
• Freshness
• Conflict of interest
• Editorial notices