docs+core: close the remaining recipe → Pattern rename gaps#1072
Conversation
A second, deeper pass beyond the phases-0-4 rename: apps/starter, README.md, and the docs/planning corpus were never swept, plus three real exported API gaps. Three genuine API gaps (missed by phase 4), each with a @deprecated alias: - parsePattern (was parseRecipe) - packages/core/src/recipes/schema.ts - destroyPattern (was destroyRecipe) - packages/dom/src/apply-recipe.ts - FIRST_RELEASE_PATTERNS (was FIRST_RELEASE_RECIPES) - packages/core/src/recipes/catalog.ts apps/starter (real, user-facing template) fully migrated: main.ts (patternById, compilePattern, applyPattern, bindData({pattern})), index.html prose + headings, and README.md (code examples + prose + the /docs/patterns URL, was a stale /docs/gallery 404). Verified at RUNTIME via a local dev server: inspect() readouts live and incrementing, the reduce/shuffle interactions genuinely re-run compilePattern() / applyPattern() / bindData() against the workspace-linked packages. README.md (root, the most-visible doc in the repo): the lane table, the FieldRecipe/ Field Formation two-lane split rewritten to the single Field Pattern lane, every code example updated to canonical names, the naming-canon spine fixed, and a genuinely broken link (/docs/gallery, no such route/redirect exists) fixed to /docs/patterns. check:readme's catalog-count check literally regexed for "N recipes" in the README -- a gate-script drift the rename itself introduced. Fixed to read FIELD_PATTERNS and look for "N patterns". docs/planning (the critical-path substrate-architecture corpus + field-bridge.md + field-os-frontier.md + docs/research/README.md): swept to Pattern, INCLUDING a real naming-collision catch -- doc 03's proposed FieldFormationReading/formations field (tracking which authored Patterns are applied) is a DIFFERENT, unshipped concept from the already-shipped types.ts formations/activeFormations/FormationChange (the setFormation MODE system). Renamed the former to Pattern language, reverted an over-eager rename of the latter (doc 02's activeFormations, which IS the real shipped field) back to Formation, and added a clarifying note so a future implementer doesn't collide the two. doc 01 (the original 3-lane "Pattern vs Field Formation vs FieldRecipe" proposal) is marked SUPERSEDED and kept as historical record, not rewritten. docs/analysis/docs-alignment-report.md (a dated completed report) is correctly left untouched, matching the established don't-rewrite-history precedent. Verified: typecheck + build + full test suite (0 fail across all 8 packages/apps) + check:api + check:docs + check:recipes + check:readme + check:links, all green.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 3331a09fce
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| ``` | ||
|
|
||
| Browse and run all 64 recipes at the [recipe gallery](https://fundamental-engine.com/docs/gallery), pick apart a | ||
| Browse and run all 64 patterns at the [pattern gallery](https://fundamental-engine.com/docs/patterns), pick apart a |
There was a problem hiding this comment.
Send pattern gallery links to
/patterns
When readers click this “Browse and run all 64 patterns” link, they land on /docs/patterns, which I checked is the Pattern model/schema page (apps/site/src/pages/docs/patterns.astro) and itself points users to /patterns for the gallery; the actual live catalog route is apps/site/src/pages/patterns/index.astro, with per-pattern pages under /patterns/${id}. This also makes the new starter link to /docs/patterns#priority-well miss the specific pattern, since that anchor is not on the model page. Use /patterns for the gallery and /patterns/priority-well for the specific pattern.
Useful? React with 👍 / 👎.
A second, deeper pass beyond phases 0–4:
apps/starter, rootREADME.md, and thedocs/planningcorpus were never swept, plus three real exported API gaps.Three genuine API gaps (missed by phase 4), each with a
@deprecatedalias:parsePattern(wasparseRecipe) —packages/core/src/recipes/schema.tsdestroyPattern(wasdestroyRecipe) —packages/dom/src/apply-recipe.tsFIRST_RELEASE_PATTERNS(wasFIRST_RELEASE_RECIPES) —packages/core/src/recipes/catalog.tsapps/starter(a real, user-facing template) fully migrated:main.ts(patternById,compilePattern,applyPattern,bindData({pattern})),index.htmlprose + headings,README.md(code examples + prose + fixed a genuinely broken/docs/gallery404 →/docs/patterns). Verified at runtime via a local dev server —inspect()readouts live and incrementing, the shuffle interaction genuinely re-runsbindData/compilePattern/applyPatternagainst the workspace-linked packages.Root
README.md(the most-visible doc in the repo): the lane table + the oldFieldRecipe/Field Formationtwo-lane split rewritten to the single Field Pattern lane, every code example updated to canonical names, the naming-canon spine fixed.check:readmeliterally regexed for"N recipes"in the README — a gate-script drift the rename itself introduced. Fixed to readFIELD_PATTERNSand look for"N patterns".docs/planning(the critical-path substrate-architecture corpus +field-bridge.md+field-os-frontier.md+docs/research/README.md) swept to Pattern — including a real naming-collision catch: doc 03's proposedFieldFormationReading/formationsfield (tracking which authored Patterns are applied) is a different, unshipped concept from the already-shippedtypes.tsformations/activeFormations/FormationChange(thesetFormationmode system). Renamed the former to Pattern language, reverted an over-eager rename in doc 02 (activeFormations, which is the real shipped field) back to Formation, and added a clarifying note. Doc 01 (the original 3-lane proposal) is marked SUPERSEDED and kept as historical record.docs/analysis/docs-alignment-report.md(a dated completed report) is correctly left untouched.Verified: typecheck + build + full test suite (0 fail across all 8 packages/apps) +
check:api+check:docs+check:recipes+check:readme+check:links, all green.