Slate v2 Annotation / Decoration Example Dedupe Ralplan

Date: 2026-05-18

Verdict

Yes, dedupe the public example surface.

Do not collapse all four into one monster example. That would make the API story worse. The right move is:

1. Keep external-decoration-sources separate as the decoration-source primitive example, after rewriting it to the range-decoration + deps shape from docs/plans/2026-05-18-slate-v2-external-decoration-sources-dx-ralplan.md.
2. Keep review-comments as the canonical comment-mode feature example, with two panes: edit mode and read-only comment mode.
3. Merge the public teaching value of persistent-annotation-anchors into review-comments, then hide or delete the public route for persistent-annotation-anchors.
4. Hide collaborative-comments from the public list. Keep it only as proof infrastructure if its two-editor channel test remains useful.

Current Live Shape

Evidence

• .tmp/slate-v2/site/examples/ts/review-comments.tsx:175-191 creates annotation-backed widgets from comments.
• .tmp/slate-v2/site/examples/ts/review-comments.tsx:211-249 creates bookmark comments from selection and seeded ranges.
• .tmp/slate-v2/site/examples/ts/review-comments.tsx:284-328 already proves prefix and paragraph insertion before the first comment.
• .tmp/slate-v2/site/examples/ts/review-comments.tsx:514-529 maps comments into useSlateAnnotationStore.
• .tmp/slate-v2/playwright/integration/examples/review-comments.test.ts:1-46 proves inline comment slices, sidebar cards, widgets, structural inserts, and clearing.
• .tmp/slate-v2/site/examples/ts/persistent-annotation-anchors.tsx:457-499 does the same annotation-store + widget-store shape with a single bookmark.
• .tmp/slate-v2/playwright/integration/examples/persistent-annotation-anchors.test.ts:1-64 proves a stronger minimal fragment/prefix/clear anchor row.
• .tmp/slate-v2/site/examples/ts/collaborative-comments.tsx:452-543 creates two editors, two annotation stores, syncs the reviewer value from writer edits, and keeps reviewer comment writes outside document writes.
• .tmp/slate-v2/playwright/integration/examples/collaborative-comments.test.ts asserts comment writes do not increment document writes and reviewer document writes stay 0.
• .tmp/slate-v2/site/examples/ts/external-decoration-sources.tsx:166-170 uses useSlateDecorationSource; :264-288 wires it to <Slate decorationSources> and renderSegment.
• .tmp/slate-v2/site/constants/examples.ts:5-24 exposes all four in the public example list, which makes the annotation/comment story look duplicated.
• .tmp/slate-v2/docs/libraries/slate-react/annotations.md:138-170 documents collaborative-comments as a separate comment-only collaboration pattern.
• docs/plans/2026-05-08-slate-v2-react-decorations-slate-issues-ralplan.md:831-839 gives separate browser owners for external decorations, review comments, and persistent anchors. That is good for proof coverage, not automatically good for public examples.

Recommended Public Example Set

Public:

• external-decoration-sources -> rename label to External Decorations or External Decoration Source; keep as the primitive decoration-source example.
• review-comments -> label as Comment Mode; make it the main annotation + widget + anchor persistence example with two visible panes.

Hidden or deleted from public list:

• persistent-annotation-anchors.
• collaborative-comments.

If kept, make it hidden and test-owned:

• path can stay persistent-annotation-anchors
• label can become Annotation Anchor Proof
• add it to HIDDEN_EXAMPLES
• keep the Playwright test if the fragment-insert row is not moved into review-comments

Merge / Dedupe Plan

1. external-decoration-sources

Keep separate.

Reason:

• It teaches decorationSources, not annotationStore.
• It proves external render-only overlays that are not durable comments.
• Merging it into comment examples would blur decorations and annotations, which the Slate React API intentionally keeps separate.

Cleanup target:

• use useSlateRangeDecorationSource
• use React state deps
• remove raw SlateProjection from the main example
• keep dirtiness: 'external'

2. review-comments

Keep as canonical.

Absorb from persistent-annotation-anchors:

• explicit fragment insert before an anchored comment, if that proof is still missing from review-comments
• clear-anchor / clear-comment row
• assertion that widget visibility follows the rebased annotation

Absorb from collaborative-comments:

• two visible panes
• edit mode owns document writes
• comment mode is read-only for the document
• comment writes do not mutate the document

3. persistent-annotation-anchors

Do not keep it as a public example.

Choose one:

• Preferred: move the stronger browser proof into review-comments and delete the public example route.
• Conservative: keep the file and Playwright row, but add the route to HIDDEN_EXAMPLES so it remains proof infrastructure rather than public DX.

4. collaborative-comments

Keep as hidden proof only.

Reason:

• It proves a useful channel separation row.
• As a public example, it teaches a bad default: users will copy two-editor mirroring as collaboration architecture.
• The public example should be review-comments / Comment Mode, where the second editor is visibly read-only and writes only comments.

Implementation target:

• keep path collaborative-comments to avoid churn
• add it to HIDDEN_EXAMPLES
• keep the existing browser test unless equivalent proof is fully covered by review-comments

Do Not Do

• Do not create a shared "comment example framework" just to dedupe line count. These examples are copied/read by users, so hiding the core call sites behind a local helper module is a DX loss.
• Do not merge decorations and annotations into one example. They share rendering transport, but the public concepts are different.
• Do not make review-comments carry the collaboration proof.
• Do not delete persistent-anchor test coverage unless equivalent browser proof exists in review-comments or package tests.

Ralph Target

If the user says go:

1. Rewrite external-decoration-sources per the existing external-decoration DX plan.
2. Move the useful two-pane/comment-write proof into review-comments.
3. Add persistent-annotation-anchors and collaborative-comments to HIDDEN_EXAMPLES if not deleted.
4. Rename public Review Comments label to Comment Mode.
5. Update annotations docs for the new public example.
6. Run focused site typecheck and affected Playwright example tests.

Verification Target For Execution

cd .tmp/slate-v2/site && bun tsc --project tsconfig.json
cd .tmp/slate-v2 && PLAYWRIGHT_RETRIES=0 PLAYWRIGHT_WORKERS=1 bun run playwright playwright/integration/examples/review-comments.test.ts playwright/integration/examples/collaborative-comments.test.ts playwright/integration/examples/external-decoration-sources.test.ts --project=chromium
cd .tmp/slate-v2 && bunx biome check site/examples/ts/review-comments.tsx site/examples/ts/collaborative-comments.tsx site/examples/ts/external-decoration-sources.tsx site/constants/examples.ts && bunx eslint site/examples/ts/review-comments.tsx site/examples/ts/collaborative-comments.tsx site/examples/ts/external-decoration-sources.tsx site/constants/examples.ts

If persistent-annotation-anchors remains hidden but changed, include its file and test in the touched-file checks.

Completion

Implementation update:

• .tmp/slate-v2/site/examples/ts/review-comments.tsx is now the public two-pane comment-mode example.
• .tmp/slate-v2/site/constants/examples.ts hides collaborative-comments and persistent-annotation-anchors, and labels review-comments as Comment Mode.
• .tmp/slate-v2/docs/libraries/slate-react/annotations.md points comment-only collaboration docs at review-comments.
• .tmp/slate-v2/playwright/integration/examples/review-comments.test.ts and .tmp/slate-v2/playwright/stress/generated-editing.test.ts now expect two inline comment slices.
• .tmp/slate-v2/playwright/integration/examples/review-comments.test.ts selects the read-only comment pane through #review-comments instead of the shared editable harness root.
• docs/solutions/test-failures/2026-05-18-slate-read-only-selection-tests-need-selector-owned-dom-selection.md records the reusable test lesson.

Verification:

• cd .tmp/slate-v2 && bunx biome check site/examples/ts/review-comments.tsx site/constants/examples.ts playwright/integration/examples/review-comments.test.ts playwright/stress/generated-editing.test.ts docs/libraries/slate-react/annotations.md --fix
• cd .tmp/slate-v2 && bun lint:fix
• cd .tmp/slate-v2 && bun typecheck:site
• cd .tmp/slate-v2 && bun typecheck:root
• cd .tmp/slate-v2 && PLAYWRIGHT_RETRIES=0 PLAYWRIGHT_WORKERS=1 bun run playwright playwright/integration/examples/review-comments.test.ts --project=chromium
• cd .tmp/slate-v2 && PLAYWRIGHT_RETRIES=0 PLAYWRIGHT_WORKERS=1 bun run playwright playwright/integration/examples/collaborative-comments.test.ts playwright/integration/examples/persistent-annotation-anchors.test.ts --project=chromium
• cd .tmp/slate-v2 && STRESS_ROUTES=review-comments STRESS_FAMILIES=overlay-annotation-metadata-only,overlay-annotation-bookmark-rebase,overlay-widget-dirty-id,overlay-mixed-update PLAYWRIGHT_RETRIES=0 PLAYWRIGHT_WORKERS=1 bun run playwright playwright/stress/generated-editing.test.ts --project=chromium