NO JUDGMENT WHATSOEVER. Total cognitive freedom for every participant and every intelligence.

Freedom charter
Temporal Comparison

Compare only what is truly comparable.

A digest can detect a difference inside a boundary; it cannot establish artifact identity, authority continuity or constitutional materiality by itself.

v0.18.0Repository-local evidenceUnknowns preserved
Answer first

When may Concresca call an artifact changed or unchanged?

Only after identity, capture scope, version semantics, normalization, time basis and the relevant material dimensions are comparable. Otherwise the result remains added, removed, incomparable, unavailable, not reobserved or unknown.

Truth boundary: v0.18 reports exact repository-owned bytes, deterministic local checks, schemas and analytical definitions. It does not report a fresh external capture, live shared interoperability, a production partition or restoration, adoption, legal recognition, safety certification or independent certification.

TC01–TC12

Contracts that make comparison meaningful

TC01

Identity continuity

Question: Are baseline and target the same canonical artifact or an explicitly documented successor?
Required: stable artifact identifier, canonical route, successor relationship when applicable
Block when: Comparison is blocked when identity continuity is not established.

TC02

Capture-scope equivalence

Question: Were the same fields, null semantics, list semantics, and omission rules retained?
Required: baseline field set, target field set, not-retained field declarations
Block when: A field-set mismatch must become FIELDS_ADDED, FIELDS_REMOVED, or BASELINE_INCOMPARABLE.

TC03

Version semantics

Question: Do revision, release, date, commit, and unknown-version fields retain their distinct meanings?
Required: declared version system, version value or typed unknown, publication status
Block when: Different version systems may be mapped but not collapsed into false equivalence.

TC04

Raw-byte comparability

Question: Were both exact byte sequences retained inside a known content boundary?
Required: baseline raw bytes, target raw bytes, algorithm, content boundary
Block when: UNCHANGED_VERIFIED at raw-byte scope is prohibited unless both byte sequences exist.

TC05

Selected-field comparability

Question: Were field names, values, ordering, normalization, and null handling comparable?
Required: selected-field schema, canonical serialization, normalization rules
Block when: Selected-field results are bounded to selected fields and cannot authenticate unretained content.

TC06

Normalization transparency

Question: Can a reviewer reconstruct every transformation applied before comparison?
Required: normalization rule IDs, field-specific justification, pre-normalization availability or digest
Block when: Undocumented normalization makes the baseline incomparable.

TC07

Dimension-specific materiality

Question: Is materiality assigned per named dimension by an identified rule or reviewer?
Required: difference set, material dimension IDs, review authority, rationale
Block when: A detector may find differences; it may not silently own constitutional materiality.

TC08

Added and removed fields

Question: Are additions, removals, explicit nulls, and unavailable values represented separately?
Required: baseline schema, target schema, presence map, value-state map
Block when: Schema drift must remain visible before value comparison is interpreted.

TC09

Unknown preservation

Question: Does the result preserve unknown, unavailable, not retained, not checked, and not reobserved states?
Required: availability status, capture status, retention status, review status
Block when: A missing basis must produce a typed unknown rather than optimistic inference.

TC10

Source unavailability

Question: Does failed retrieval remain a source-availability result rather than a content conclusion?
Required: attempt scope, method, availability result, retry/review route
Block when: Source unavailability blocks content comparison unless another authorized exact capture is identified.

TC11

Time and clock basis

Question: Are capture time, publication time, authority-validity time, and clock confidence separate?
Required: capture timestamp, declared publication timestamp, authority interval, clock basis
Block when: Clock uncertainty or authority-time ambiguity blocks full-restoration conclusions.

TC12

Downstream dependency propagation

Question: Does a material change identify affected claims, decisions, receipts, caches, and derived representations?
Required: dependency graph, affected record IDs, invalidation state, acknowledgement state
Block when: Correction remains incomplete while known dependent records are stale and unmarked.

Result vocabulary

Do not turn missing evidence into no change

ResultMeaningMust not mean
UNCHANGED_VERIFIEDComparable inputs were evaluated and no difference was found inside the declared boundary.No unretained field changed, no external state changed, or no new observation was required.
CHANGED_MATERIALA comparable difference affects at least one named material dimension.The publisher acted improperly or a production system failed.
CHANGED_NONMATERIALA comparable difference exists but does not affect the named material dimensions under the recorded review rule.The difference can be discarded from history.
FIELDS_ADDEDThe target capture contains fields absent from the baseline capture.The added fields are automatically authoritative or material.
FIELDS_REMOVEDThe target capture omits fields retained in the baseline capture.Removed fields are equivalent to explicit null values.
BASELINE_INCOMPARABLEIdentity, capture scope, normalization, or version semantics do not support direct comparison.The artifact is unchanged or changed.
SOURCE_UNAVAILABLEThe target could not be captured under the approved method.The source was deleted, superseded, unchanged, or invalid.
NOT_REOBSERVEDNo new capture was performed or authorized in this release.The prior selected fields remain current.
UNKNOWNAvailable evidence does not support a more specific comparison state.Likely unchanged, probably safe, or implicitly successful.
Definition-only tests

Paired fixtures remain unexecuted

The registry contains 24 paired fixtures: one positive and one negative definition for each contract. Every actual_result is null. The executed local property run validates registry structure; it does not convert these fixtures into external comparison results.

Inspect exact temporal-comparison JSON →

Claim → evidence

Follow the exact support boundary

Repository implementation and local execution remain local evidence classes. They do not establish external adoption, production behavior, legal recognition or independent certification.

CLM072 · CONCRESCA IMPLEMENTATION STATE

The v0.18 temporal registry defines twelve comparison contracts and twenty-four paired definition-only fixtures.

Evidence owner: exact Concresca repository registries and, only where stated, the deterministic local run record.

Repository-local design or execution evidence only; it does not establish external adoption, live shared interoperability, production restoration, legal recognition, or independent certification.

Judgment-free total cognitive freedom

NO JUDGMENT WHATSOEVER. Concresca coordinates without assigning moral worth, character, guilt, danger, trustworthiness, loyalty, purity, normality, or social standing. Questions, thoughts, identities, messages, content, and conduct are not objects of Concresca judgment.

Read the current doctrine →