Migrating to v11 without destroying historical state
v11 is additive: preserving valid v9 logical state and v10 hosting remains the default. A v10 hosted node does not need its v9 state or published bytes rewritten merely because v11 exists.
Choosing a path
Path A — Existing legacy Kristal: leave its historic contract intact. Inventory source identity strings and older labels; use Legacy-Migration-v11 only when adopting a strict v11 semantic graph is explicitly desired. Build new logical artifacts/revisions, never mutate the old committed ones.
Path B — New strict-v11 Kristal: allocate namespace-scoped KQ/KP/KA/KS with the internal authority; create a graph, dictionaries and Wikidata crosswalk, register artifacts in Authoring Kit 4 and enable v11 checks. Q… never replaces KQ… in the graph itself.
Path C — Arketype design scaffold: create --authoring v11 yields an empty Kit 4 workspace and readiness guide, not a compliant graph. Finish authoring and register real artifacts before opting into strict conformance.
Source audit and migration gates
- Pin original source hashes, commitments and the old schema/version.
- Distinguish legacy semantic identifiers from textual/literal content.
- Generate candidate KQ/KP resolution proposals; review ambiguity and provenance.
- Publish reviewed crosswalk and dictionary layers as separate immutable versions.
- Only after approval produce a new canonical graph and Logical Artifact commitments.
- Run the Framework TCK and Kit-specific tests; preserve original snapshots and explicit rollback route.
An absent Wikidata Q or a missing translation does not invalidate an existing internal KQ. The same graph can be rendered in many languages as dictionaries evolve without creating new semantic identities.
See also: Compatibility-v9-v10 · Conformance-v11 · Wikidata-Crosswalk-v11.