Source archive

A retained-document archive. Each record fixes one URL at one moment — the URL, the date it was requested, a SHA-256 of the file as retrieved where a file was retrieved, and, where this record holds one, the address of a third-party capture. Records are written one at a time, when a document is captured or when a link recorded for a source is found dead: 18 records across 18 sources.

Documents are linked, never rehosted. The hash is the integrity claim: it lets anyone confirm that the file they hold is the file that was read on that date.

Observation, retained snapshot, third-party capture

Observation

One dated fetch of a tracked source, made by this record. Its day is this record's, and its content_hash, where one was produced, is over what this record fetched. 24,082 are recorded, published at /api/v1/observations.json.

Retained snapshot

What this record kept from a retrieval of its own: a retained record or a retained reading. A retained record's SHA-256, where a file was retrieved, is over that one file as retrieved; of the 18 records listed below, 13 carry one, and for the other 5 no file was retrieved and no SHA-256 is recorded. A retained reading's digest is over normalised text. Every digest of either kind was computed by this record. The readings are under retained readings.

Third-party capture

An address minted at a web archive that names a publisher's page. This record holds the address only: it has not read what the address serves, holds no digest of it and serves no copy of it. Its day is the archive's own stamp, read from the address. A third-party capture address is not an observation and not a retained snapshot. The addresses are counted under third-party captures.

Third-party captures

Retained records holding the address of a third-party capture: 1 of the 18 above, in the column of that name. Separately, observations of 119 of the 132 tracked entries carry such an address: 204 capture addresses in all, each counted once per entry by the archive's own address. A route to 139 of them is published and 65 are withheld, decided per publisher host from the terms filings this record holds. Every address this record holds is at web.archive.org.

Each tracked entry's captures, with the archive's stamps and the reason for every withheld route, are in that entry's TimeMap, indexed at /api/v1/timemap.json. An address in an observation is what the archive answered when this record asked it to save a page; it is evidence that a request was answered, and this record holds no reading of the capture behind it.

Retained readings, and comparing two of them

Separately from the records above, this site retrieves each tracked source on a repeating schedule and, where text can be extracted, retains the normalised text it hashed. Where two of those readings differ, the difference between them can be computed.

Two or more distinct retained readings are held for 90 of the 132 tracked entries. Every pair of them carries either the difference or the condition that refused it, and the conditions are stated at the comparison index.

Last updated: 2 November 2026