Komorebi Methodology: proofreading and HTML reading edition

Reviewed 2 November 2026 from collection revision 75caee1d5ebc01d64b439101d94cd5d1c1fc3bb0.

Proofreading

The review covered the full preface, all nine numbered essay chapters, the closing reflection, source and verification notes, nine bibliography entries, three figure captions, and local notation and cross-references. The sources are Komorebi_Methodology.tex, essay.tex, and sources.tex in monographs/komorebi-methodology/.

Routine edits: none. No objective spelling, grammar, punctuation, or LaTeX error requiring a correction was identified in this pass.
Mathematical-token changes: none.
Unresolved proofreading issues: none identified within that scope.
The manuscript, original diagrams, PDF, delivery manifest, and historical audit files are unchanged. This pass did not audit the mathematical proofs, verify external citations, execute Lean, or repeat the historical PDF layout review.

The existing source package already records its written review and 28-page PDF review. Those are inherited records, not new review claims from this conversion.

Reading conversion

The edition contains 12 source-derived text pages plus references and contents. The authored order is preface, nine chapters, closing reflection, then source and verification notes. The print cover and automatically generated printed contents are represented by the HTML title and chapter index.

The conversion preserves all 164 mathematical expressions in order, three global equation numbers, unnumbered subsection headings, nine bibliography entries, and three figure captions. The browser diagrams are byte-identical copies of the PNGs already supplied with the source package, with receipts binding each to its original PDF and PNG. The published source downloads retain their exact bytes.

Print-only wrappers are translated into the existing reader panels and headings. The superscript source citations become numbered bibliography links. Source-line anchors, previous/next links, chapter filtering, text settings, saved passages, comparison links, search entries, and the work overview use the shared reader.

tools/methodology_sections.json records exact source digests and contiguous ranges. Changed sources stop the build pending a reviewed selection. The source package's byte-preservation Git attribute prevents Windows newline conversion from invalidating its existing delivery manifest. The rule lives in monographs/.gitattributes, preserving the root attributes file's identity in the frozen Euler example record.

Validation

python -B -m unittest discover -s tools -p 'test_methodology_reading.py' -v: all three checks passed, including the normal published-equation checker, equation and figure numbers, citation and passage targets, source drift rejection, heading structure, and original image identities.
Shared navigation, reader display, and usability checks: 26 tests passed.
python -B verify_manifest.py: all 54 payload files matched the delivered manifest (55 package files including the manifest).
In an isolated copy, exact_checks.py: 40/40 groups passed, 665 finite cases; integrity_checks.py: 17/17 controls passed. The retained package and its historical receipts were not overwritten. Finite cases do not establish the unbounded arguments.
Browser inspection confirmed chapter navigation, rendered equations, equation labels, the memory derivation, and the chapter contents panel. Browser checks inspect presentation, not mathematical correctness.

The website workflow includes the new source-preservation tests and runs its complete site and link checks before publication.

The first full integration run exposed a test assumption that every HTML edition must have Lean sources. The methodology has no selected Lean resource and makes no Lean claim. The corrected test still requires receipt-owned source selections for every work declaring Lean resources, verifies that every reading edition has a component record, and checks that an edition without a Lean selection acquires no inspector or implied coverage.