Skip to content

Chemistry evidence-watch baseline review — 1 October 2026

Decision and scope

This is a maintenance acknowledgement of Chemistry goal wording already committed in PR #68, merged as cbc32044f55223979815dc0eef3757caee82a5e0. The PR's Chemistry evidence-watch job failed with changed=1 added=0 removed=0; its uploaded delta identifies only the canonical Chemistry JSON. This review does not edit canonical content, reapprove learning goals, claim human approval, or complete Chemistry M7.

The prior baseline at 2026-09-07T08:01:17Z hashes the canonical source at the first parent of the merge (cbc32044f^, also 4eb105ac1) to:

cc1d63eed31a93420a2dd34890aed18bd62119fc42d50420e55ba9ee6e5a4ba3

The same canonical-evidence-json-v1 normalization of the merged canonical source hashes to:

64e40c82adb83f37a397e512de0ee645658a3ea3e1802e51baba124bcb1aa8d5

The normalization ignores only goal-visualization resource links and JSON formatting. It continues to watch descriptions, graph structure, source links, and other canonical fields.

Exact semantic delta

A field-by-field comparison of those two normalized sources finds 473 identical goal IDs and exactly the following ten changed fields. There are no added or removed goals, top-level changes, or other goal-field changes. The full previous and current wording is in git diff cbc32044f^ cbc32044f -- curricula/DE/Gymnasium/canonical/DE_DEU_S_GYM_CANONICAL_CHEMIE.de.json.

Goal ID Changed fields Reviewed substance of the new wording Existing machine-review evidence
326d45bf-9f77-57d5-a054-93e76b034dd5 description, descriptionEn Distinguishes solid, liquid and gas by observable properties; treats transitions as states of the same substance. chemie/rollout-v1/2026-09-30/batch-004w-states-corrected-image-current-1-v1 in curricula/DE/Gymnasium/quality/goal-description-review/
d2ccd1d5-56f7-583f-9724-e97441367f91 description, descriptionEn Uses the actual indicator colour scale and broad pH ranges for suitable school experiments. chemie/rollout-v1/2026-09-30/batch-004v-indicator-reference-current-1-v2 in the same review root
8d4ef102-e6a6-4d2e-bb6b-e707d3f2e566 description, descriptionEn Uses formation of new substances to distinguish reactions; temperature or colour changes alone are inconclusive. chemie/rollout-v1/2026-10-01/b009-reaction-vs-physical-current-1-v1 in the same review root
4dab7d52-5b89-52ab-a425-17a7707f44c8 description, descriptionEn Compares direct CO₂ release and combustion heat on consistent reference bases and system boundaries. chemie/rollout-v1/2026-10-01/b009-fuel-comparison-current-1-20261001-v3 in the same review root
0503c975-3934-5206-962e-b1c247de0c12 description, descriptionEn Explains the fossil-store-to-atmosphere transfer and atmospheric increase when total inflow exceeds total uptake. chemie/rollout-v1/2026-10-01/b009-fossil-carbon-current-1-20261001-v2 in the same review root

The B009 package's current five-gate checkpoint documents its three newly closed goals as 58/376 Chemistry goals strictly complete, with the separate human gate still open. The two B004 packages have current dual-round resolutions. This watch acknowledgement adds zero new strict goal closures.

Why the P6 watch baseline may advance

The watch manifest reopens P6 maintenance only if a canonical or tracker delta changes visible scope, source-backed projection status, runtime mapping coverage, or composition-view eligibility. The exact normalized delta above changes none of these. The checker reports the other 136 watched files unchanged, including source extractions, mappings, composition views and the rollout tracker. Their baseline records remain byte-for-byte identical. The five descriptions were changed and reviewed in the separate machine Curriculum-QS lane; this baseline change records that active source state without presenting a hash refresh as a new fachliche Prüfung.

Only the canonical Chemistry baseline record's evidence hash and observed file modification time, plus the baseline updatedAt, advance. The historical 7 September review and Git history preserve the earlier baseline. The watch manifest, normalized hash algorithm, checker, workflow and canonical source remain unchanged.

Verification

  • Compare the normalized first-parent and merged sources: exactly five goal IDs, each with only description and descriptionEn changed; all 473 IDs retained.
  • Compare watch records before and after this acknowledgement: all 136 other records are unchanged.
  • Run python3 -B scripts/canonical_chemistry_evidence_watch.py self-test and ./scripts/run_canonical_chemistry_evidence_watch.sh; the expected delta is changed=0 added=0 removed=0.