Skip to content
Property · Format description concepts

Effective claim date: when a statement in a format record applies

Published Formatory editorial team

An effective claim date is the date from which a particular statement in a registry record is asserted to hold, independent of when the statement was recorded.

Registries attach it to individual claims, such as a release, a media type registration or a change in disclosure, so that a record can say both what is true and since when.

A registry learns about a format in pieces and out of order. A specification published in 2008 may be cataloged in 2011, corrected in 2014 and linked to a new version in 2019. If the record keeps only one date, “last updated”, most of this history disappears. The effective claim date keeps a separate answer to a different question: from when is this particular statement true?

Two clocks for every statement

Database designers describe the problem as two kinds of time. Valid time is when something is true in the world. Transaction time is when the system learned or recorded it. A database that tracks both is called bitemporal, and the same distinction applies to registry metadata.

In format records the two clocks typically map like this:

Question Kind of time Typical field
When did this become true about the format? valid time effective claim date
When did the registry record it? transaction time date submitted, change notes
When was the record last touched? transaction time modification date

Keeping both is what allows a registry to tell apart what the record said in 2012 from what was true about the format in 2012, two separate questions. They often have different answers.

What gets an effective date

Not every statement needs one, and over-dating clutters a record without making it more accurate. A name or a description of structure is either right or wrong. Dates earn their place on claims that describe events or states that start at a point:

  • Release and publication: a format version first released, a specification published.
  • Registration: a media type registered with IANA, a FourCC registered with an authority.
  • Changes in status: a format’s disclosure level going from partial to full, a version declared obsolete, a patent expiring.
  • Support: software beginning or ending support for a version.

A single format record can easily hold claims with effective dates decades apart. A spreadsheet format might show a first release in the late 1980s, a media type registration much later, and an end of vendor support later still, with all three recorded in one cataloging session.

Each of these is usually backed by a source, and the reference purpose of that source tells the reader whether the date comes from the specification body itself or from a secondary account.

Precision and uncertainty

Format history is full of half-known dates. Vendors rarely announced exact release days for file format versions, and the first documented appearance of a format is often only a year. The effective claim date should carry exactly the precision of the evidence. ISO 8601 allows reduced precision (2008, 2008-02), and the Extended Date/Time Format (EDTF), now part of ISO 8601-2, adds ways to say “approximately” or “uncertain”, which are much better than an invented first of January. Generic dc:date values have the same issue, but there the date is about a whole resource, not about one claim inside it.

Changes, history and corrections

When a claim is replaced, the old claim with its effective date is history, not an error. A good registry keeps it, so that the record can show that a format was partially disclosed from 1997 to 2008 and fully disclosed afterwards. Corrections are different: a claim that was never true should be marked as corrected, with a note explaining why, the territory of skos:changeNote and history notes. Confusing a change in the world with a correction of the record is the most common mistake in dated metadata, and the two-clock model exists mainly to prevent it. The vocabulary index lists the related date and note properties, and the file format identification guide shows where version dates matter in identification.

Example in RDF

<format rdf:about="https://example.org/formats/example-office-binary">
  <claim>
    <property>disclosure</property>
    <value>Full disclosure</value>
    <effectiveClaimDate>2008-02-15</effectiveClaimDate>
    <dateSubmitted>2011-06-03</dateSubmitted>
    <source rdf:resource="https://example.org/refs/vendor-spec-v1"/>
  </claim>
</format>

Questions about effectiveClaimDate

What is an effective claim date?

It is the date from which a statement in a record is said to be true. If a specification was published in 2008 and a registry recorded that fact in 2011, the effective date of the claim is 2008 and the recording date is 2011.

How is it different from a date submitted?

The submission date says when a statement entered the registry, which is a fact about the registry. The effective date says when the statement became true in the world, which is a fact about the format. Both are worth keeping.

Why date individual claims and not the whole record?

Because a format record collects facts that change at different times. The release date, the media type registration, a new signature and a change in disclosure level each have their own date, and a single 'last modified' value hides all of them.

What if the exact date is not known?

Record the precision that is known, such as a year or a year and month, using a format like ISO 8601 or EDTF that can express reduced precision. A guessed exact day suggests a certainty the source does not support.

Can an effective claim date be in the future?

Yes. A standard can be approved with a later date of entry into force, and a vendor can announce the end of support for a format version in advance. A future effective date lets the registry record the claim now.

More from Format description concepts

All 16 in this group