Auto-save subsystem for pending suggestions by adamsilverstein · Pull Request #78308 · WordPress/gutenberg

adamsilverstein · 2026-05-14T17:53:53Z

Overview

Follow-up to #77407 (Phase 5). Carves the auto-save subsystem out of Phase 5 so each PR concentrates on one concern — Phase 5 stays focused on REST permissions + PHP tests, and this PR owns the Google Docs–style background save.

Tracked in #73411.

What's in this PR

SuggestionAutoSave — invisible component that debounces overlay edits (1.5 s per block) and persists them as note comments with the _wp_suggestion meta payload. Replaces the explicit SuggestionCommitBar toolbar button (restored on Phase 5 so the prior stack stays functional in isolation).
Per-block save queue — saves on the same clientId chain sequentially so a slow network call can't race with a follow-up edit and produce duplicate POSTs or out-of-order writes. Different blocks run concurrently.
updateSuggestion / deleteSuggestion on the provider — once a comment id is linked, subsequent edits PUT the same comment; a fully reverted overlay trashes the note. Includes a stale-link check via core-data so a peer-resolved comment doesn't get clobbered mid-session.
Overlay plumbing — commentId and syncedOpsKey fields on overlay entries, SET_COMMENT_ID / SET_SYNCED_OPS_KEY reducer cases, and matching setCommentId / setSyncedOpsKey callbacks on the context.
Tests — `test/auto-save.js` covers debounce, queue ordering, create/update/delete branching, and stale-link orphan handling. `test/overlay-context.js` covers the new reducer cases.

Test plan

Enter Suggest mode and edit a paragraph.
After ~1.5 s the network shows a `POST /wp/v2/comments` creating the note.
Continue typing — subsequent saves are `PUT`s against the same comment id.
Revert the overlay to baseline — the note is trashed.
Have a peer accept the suggestion mid-session — the next local edit creates a fresh comment instead of clobbering the resolved one.

```bash
npm run test:unit -- packages/editor/src/components/suggestion-mode/
```

Stack

🗺️ PR Stack Navigation

#	PR	Phase
1	#77403 Intent scaffolding	Edit / Suggest / View mode
2	#77404 Overlay capture	In-memory suggestion overlay
3	#77405 Provider + Accept/Reject	`_wp_suggestion` meta, provider, sidebar actions
4	#77406 Summary + docs + attribute tests	Add/Delete/Formatting summary, architecture stub, conflict scoping
5a	#78351 REST permissions and PHP coverage	Permissions, payload cap, PHP tests
5b	#78352 Summary + attribute conflict + docs	Renderer, per-attribute staleness, architecture docs
5c	#78353 Surface Apply/Reject in the collaboration sidebar	Icon buttons + e2e + sidebar wiring
6	#78308 Auto-save subsystem ← this PR	Background debounced save (replaces commit-bar)

📋 Tracking issue: #73411

When clicking "Add Note" on a block that already has an existing note, the user was being redirected to the reply field of the existing note. This prevented users from creating multiple independent note threads on the same block. This change: - Adds an `addNewNote` parameter to `openTheSidebar()` to differentiate between clicking "Add Note" (should always create new thread) and clicking the avatar indicator (should focus existing thread) - Updates `AddCommentMenuItem` to pass `{ addNewNote: true }` - Adds E2E tests to verify multiple notes can be added to the same block Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

@Mamaduka

getNoteIdsFromMetadata is a trivial pure function that normalizes a scalar/array value. The overhead of useMemo exceeds the cost of calling it directly each render. Addresses review feedback from @Mamaduka. Generated with [Claude Code](https://claude.ai/code) via [Happy](https://happy.engineering) Co-Authored-By: Claude <noreply@anthropic.com> Co-Authored-By: Happy <yesreply@happy.engineering>

@Mamaduka

Use direct property access for noteId instead of extracting metadata into a separate variable. Addresses review feedback from @Mamaduka. Generated with [Claude Code](https://claude.ai/code) via [Happy](https://happy.engineering) Co-Authored-By: Claude <noreply@anthropic.com> Co-Authored-By: Happy <yesreply@happy.engineering>

@Mamaduka

Clarify why the blur handler needs to bail out during submission: clicking the submit button triggers blur before the async onSubmit completes, which would otherwise close the form prematurely. Addresses review feedback from @Mamaduka. Generated with [Claude Code](https://claude.ai/code) via [Happy](https://happy.engineering) Co-Authored-By: Claude <noreply@anthropic.com> Co-Authored-By: Happy <yesreply@happy.engineering>

This reverts commit 7d2ff81.

Prevent blur from closing the add-note form while the async submit is in progress by guarding with isSubmittingRef. Also stop the auto-select effect from overwriting the "new" note selection, and use a ref for noteThreads so the effect only runs when blockNoteIds changes, not on every thread update.

The first "Note added." snackbar persists when adding a second note, causing a strict mode violation (2 elements) and a premature metadata assertion. Dismissing the first snackbar before adding the second note fixes both tests.

Extend E2E tests with scenarios for deleting one of multiple notes, resolving individual notes, auto-selecting the first unresolved note, and verifying metadata cleanup when all notes are removed. Add unit test edge cases for falsy noteId values, string coercion, and null/undefined metadata handling.

Reconcile the multiple-notes-per-block data model with trunk's notes architecture refactor. Keep getNoteIdsFromMetadata / addNoteIdToMetadata / removeNoteIdFromMetadata utilities and adapt useBlockComments to iterate arrays of note IDs through trunk's threadsById partition. Union the utility tests with trunk's calculateNotePositions coverage.

Resolve conflicts from PR #77614 (Notes refactor) by porting the multiple-notes-per-block feature into trunk's split component structure (notes.js, add-note.js, hooks.js, utils.js).

addNoteIdToMetadata and removeNoteIdFromMetadata now compare ids using String() so a string-typed legacy id ('5') and a numeric id (5) match. Without this, the dedupe check could miss and the array could contain duplicates that round-trip differently than expected. Adds unit coverage for: - mixed-type dedupe in addNoteIdToMetadata - mixed-type removal in removeNoteIdFromMetadata - calculateNotePositions with two and three threads sharing a block (identical blockRect.top), exercising the same-block layout path introduced by multi-note support.

Adds two e2e tests inside the 'Multiple notes per block' describe: - A legacy scalar noteId is converted to an array on first multi-note write, and the original (orphan) id is preserved as the first entry. - A block with two notes round-trips through saveDraft + page.reload with both notes still attached and the metadata array intact in the same order. Closes the two highest-leverage e2e gaps for multi-note support: the backwards-compat claim from the PR description, and end-to-end serialize/parse persistence.

The 'Multiple notes per block' describe was a sibling of 'Block Notes', not a child, so its inner test.use/beforeEach/afterAll ran in addition to the file-level ones — every test in that describe was creating each post twice and deleting all notes twice. Removes the duplicate hooks. Adds two helpers on BlockNoteUtils to collapse repeated boilerplate: - dismissSnackbar(text) for clearing the previous toast before asserting a new one - addAnotherNoteToCurrentBlock(content) for the click-add-note + fill + submit sequence used by every multi-note scenario Also adds an inline comment in useNoteActions.onCreate marking the read-modify-write race that PR #75147 alone cannot fix; a real merge strategy is tracked in #74751.

The merge from trunk (PR #77614 refactor) kept the branch's pattern of calling focusNoteThread() directly after selectNote(noteId). That call fires before the sidebar's useEffect re-renders the AddNote form, so sidebarRef.current may not yet contain the new treeitem when the MutationObserver inside focusNoteThread starts watching, and the synchronous menu close steals focus before the async .focus() resolves. Switch to selectNote(noteId, { focus: true }) so the existing useEffect in notes.js drives focus after the sidebar has rendered, matching trunk's working behaviour. Drops the now-unused focusNoteThread import.

Commit a325780 removed the per-test admin.createNewPost() and afterAll cleanup from the 'Multiple notes per block' describe on the premise they duplicated the file-level hooks. They didn't: the 'Block Notes' describe is a sibling, not a parent, so its hooks never applied to these tests. Removing them left every test in the block running against whatever page the previous run left behind, which manifested in CI as page.waitForFunction timeouts on window.wp.blocks because there was no editor page to test against. Also fix the save/reload test to explicitly open the notes sidebar after page.reload() — the pinned sidebar isn't restored on reload, so the note threads aren't in the rendered DOM until it's reopened.

Restore the selectBlock(clientId, null) call inside openTheSidebar so the List View entry point — where the block isn't already the canvas selection — keeps working. The accompanying refactor that introduced multiple-note support had dropped this in favor of the trimmed function signature. Also drop the rawNoteId rename and the redundant '?? null' coalesce on the noteId selector, per review feedback.

The 'Multiple notes per block' refactor replaced the document.hasFocus check with the isSubmittingRef guard, but they cover different cases — window/tab losing focus vs. an in-flight async submit. Keep both so neither path can prematurely close the form.

Replace the selectedNoteRef/notesRef render-time assignments with a useMemo that picks the priority thread (first unresolved, else first) and a thinner effect that bails on the 'new' state or when the user already has a thread on the block selected. This keeps explicit thread choices stable while letting status changes promote a new priority, without writing to refs during render.

Move the new describe block so it sits alongside Keyboard inside the top-level Block Notes describe and inherits the parent's beforeEach post-creation and afterAll comment cleanup, instead of duplicating them at the top level.

`AddNoteMenuItem` is a slot fill rendered per block in the List View row menus and passes the row's clientId to its onClick handler. The prior simplification of openTheSidebar's signature dropped that arg on the floor, which would target the canvas selection instead of the block whose menu was opened. Accept an optional clientId option on openTheSidebar, look up threads for that block directly so the right note is resolved, and pipe the slot-provided clientId through from AddNoteMenuItem.

The 2709799 refactor moved selectedNote into the auto-select effect's deps, so collapse paths (Escape/Cancel/Shift-Tab) immediately re-fired the effect and re-expanded the just-collapsed note. Track selectedNote in a ref updated inside an effect — keeps Mamaduka's no-refs-during- render rule and runs the auto-select only when the block context actually changes.

Links to wordpress-develop#12215, which backports the render_block filter that strips inline note markers from rendered output.

Inline-note highlights are rendered through the annotations API: each note adds a `core-note` annotation, which the editor paints as a `#annotation-text-{id}` mark. Switching to the code editor and back left these highlights blank until the next block edit or a reload. Two things conspire: - Entering the code editor unmounts the canvas RichText, whose `core/annotation` change handler removes the note annotations from the store. Returning to the visual editor does not re-parse the blocks, so `useAnnotateBlocks`'s memoized `annotations` keeps the same reference and the add-once effect never re-runs. - The notes sidebar unmounts in code view, unregistering the `core/note` format. On remount the `annotations` memo runs once while the format is still missing, so `findNoteInBlock` can't resolve any markers and caches an empty result that never recomputes. Track the `core/note` format registration reactively so the derivation re-runs once it is registered again, and reconcile the effect against the annotations actually present in the store (add missing, remove stale) instead of adding once. The highlights now survive repeated code-editor round-trips. Add an e2e regression test covering the round-trip.

adamsilverstein · 2026-06-18T12:15:08Z

Thanks @saroshaga - both are real and now fixed in 39ac4ff.

1. Duplicate note while a save is in flight. You were right: the per-block queue serialized execution, but the new comment id flowed back only through React state (setCommentId) and was read via a ref refreshed in an effect. A save chained behind an in-flight create runs as a microtask the moment the create resolves - before React commits the id - so it read null and POSTed a second note. Fixed by tracking the id (and the synced-ops fingerprint) in synchronous ref maps written the instant a save learns them, so a queued save always sees the latest linkage regardless of render timing. The maps are seeded from overlay state on first touch (remount-safe) and pruned when an overlay entry clears so a reused clientId can't inherit a stale id.

2. Updating a peer-resolved note from stale cache. Also correct - the check read the comment from this tab's core-data cache, which can pre-date a peer's accept/reject. Now we force a fresh read before updating. One subtlety worth flagging: a single-item GET /comments/<id>?context=edit requires moderate_comments, which a suggester usually lacks, so it would 403 and silently fall back to stale cache. Instead it queries the note collection filtered to the id (?post=…&type=note&include[]=…&context=edit) - the path the note REST controller grants to edit_post users, same as the sidebar - and mirrors the result back into core-data.

Both scenarios have regression tests (test/auto-save.js), each verified to fail without its fix.

The suggest-mode stack is meant to ship on top of the inline-notes feature (#78218): the stack's e2e specs exercise the inline "Add note" rich-text format and per-author highlight tint, but the implementation (the `core/note` format, `note-highlight-styles`, and the annotate/format-registration wiring) was never merged in, so those specs failed. Merge add/inline-notes-hybrid so the inline-notes layer is a true ancestor. Conflict resolutions: - collab-sidebar/note.js: keep inline-notes' collapsible "Show more" content and re-apply the suggest-mode SuggestionActions integration (hide Resolve on suggestion notes, surface Accept/Reject). - block-notes.spec.js: take inline-notes' canonical inline-note specs. - lib/compat PHP: keep the suggest-mode suggestion backend (the `_wp_suggestion` meta + REST suggestion-lifecycle permissions); inline-notes only relocated the inline-marker stripping. - CHANGELOG / suppressions: union.

Per-block debounced background save for Suggest mode. After a 1.5 s idle window, pending overlay edits persist as a note comment; subsequent edits on the same block update the existing note's meta rather than creating a new one, and a fully reverted overlay trashes the note. Saves on the same clientId run sequentially via a per-block promise queue, while different blocks run concurrently. Replaces the explicit SuggestionCommitBar toolbar button.

Move the auto-save subsystem's ref syncing out of render and into an effect, matching the store-interceptor fix, so the react-hooks rule against accessing refs during render passes. Prune the now-unused auto-save.js suppression along with the stale bulk-suppressions the stack carried before trunk's ESLint dependency bump. lint:js passes.

…ution races Address two timing concerns raised in review of the auto-save subsystem. 1. Duplicate note when the comment id has not propagated yet. The per-block queue serialized save execution, but the new comment id flowed back only through React state (setCommentId) and was read via entriesRef, which is refreshed in an effect. A save chained behind an in-flight create runs as a microtask the instant the create resolves, before React commits the id, so it read null and POSTed a second note. Track the id (and the synced-ops fingerprint) in synchronous ref maps written the moment a save learns them, so a queued save always sees the latest linkage regardless of render timing. The maps are seeded from overlay state on first touch (remount-safe) and pruned when an overlay entry is cleared so a reused clientId can't inherit a stale id. 2. Updating a note a peer already resolved. The stale-link check read the comment from this tab's core-data cache, which can pre-date a peer's accept/reject, so a resolved note could be clobbered. Force a fresh read before updating, querying the note collection filtered to the id (the path the note REST controller grants to edit_post users; the single-item edit context requires moderate_comments) and mirror the result back into core-data so the sidebar stays in sync. Add regression tests for both, each verified to fail without its fix.

Three e2e tests asserted behavior not present on this branch; they were masked until the JS build break was fixed and the suite could run. editor-intent-switcher: the intent is session-scoped by design (see setEditorIntent / editorIntent reducer) - reload returns to Edit. Rewrite the reload test to assert that reset instead of expecting Suggest to persist, which the implementation deliberately refuses. suggestion-mode: 'auto-saves a content edit' and the heading-level capture test wait for an automatic debounced POST /wp/v2/comments. The debounced auto-save subsystem is Phase 6 (#78308), not in this stack's ancestry; here a suggestion is only persisted via the explicit commit-bar Submit button. Skip both with a tracking reference to #78308.

Bring the auto-save branch up to date with its base after the stack was cascaded onto latest trunk. Reconcile the divergent REST layer onto phase-5b's architecture: phase-5b defers note permissions to core (Notes GA in 6.9) and layers only the suggestion-specific behavior via Gutenberg_REST_Comment_Suggestions_Controller (lib/compat/wordpress-7.1). The older monolithic class-gutenberg-rest-comment-controller-6-9.php carried by autosave duplicated core's note-permission handling and was already dead code (never required), so it is removed; the suggestion REST test now targets the 7.1 controller. Keep autosave's JS auto-save subsystem (it replaces the commit bar): resolve the suggestion-mode component conflicts to the auto-save versions and drop the orphaned commit-bar.js re-introduced by the merge. store-interceptor.js keeps the effect-synced refs (react-hooks compliant). Regenerate eslint suppressions and refresh the architecture doc's REST/PHP table.

…ta is registered The note/suggestion comment meta (`_wp_note_status`, `_wp_suggestion`, `_wp_suggestion_status`) is registered by `gutenberg_register_block_comment_metadata()` in `lib/compat/wordpress-6.9/block-comments.php`, but `lib/load.php` only required the 7.1 inline-notes companion — never the 6.9 base file. As a result the function was undefined at runtime, so the REST suggestion meta was never registered and the `WP_Test_REST_Comments_Controller_Gutenberg` suite errored on every test (`Call to undefined function gutenberg_register_block_comment_metadata()`) in its `set_up()` re-registration hook. Require the 6.9 file before the 7.1 companion. The file self-hooks on `init` and guards its helpers with `function_exists`, so it composes cleanly with core's GA notes. Verified: full PHP suite green (1991 tests, 0 failures) with this require in place.

…ments test `ReflectionMethod::setAccessible( true )` is deprecated as of PHP 8.5 (it has been a no-op since 8.1, which grants reflective access automatically). PHPUnit promotes the deprecation to an error, so `test_lifecycle_update_rejects_non_allowlisted_fields` and `test_lifecycle_update_accepts_form_encoded_bodies` errored on the PHP 8.5 CI job while 7.4 passed. Remove the two no-op calls; the immediately following `->invoke( null, $request )` still works on 8.1+. Verified green on PHP 8.5.7 (wp-env).

Removing setAccessible() outright fixed PHP 8.5 (where it is a deprecated no-op) but broke PHP 7.4, which still requires it to invoke the private static is_suggestion_lifecycle_update() via reflection. Call setAccessible() only on PHP < 8.1, where it is both required and not deprecated; on 8.1+ reflective access is automatic and the call is skipped to avoid the 8.5 deprecation. Verified the pattern on PHP 7.4.33 and 8.5 (both invoke cleanly, no deprecation); PHPCS clean.

The `static-checks` CI job (lint:js) failed with "There are suppressions left that do not occur anymore" — the trunk cascade (and the suggestion-mode code changes) eliminated violations whose entries still lingered in `tools/eslint/suppressions.json`. Ran `npm run lint:js:prune-suppressions` to drop the 92 now-unused entries. Verified `lint:js`, `lint:css`, type-check build, and `other:check-local-changes` all pass.

… into suggest-mode-autosave

… into suggest-mode-autosave # Conflicts: # phpunit/experimental/class-wp-rest-comments-controller-gutenberg-test.php # tools/eslint/suppressions.json

The phase-5a fix added a wordpress-6.9/block-comments.php require, which the merge into autosave duplicated against the one this branch already had. Two plain require() calls of the same file trigger a redeclare fatal that breaks PHP boot and every Playwright shard. Drop the duplicate so the file loads once.

github-actions · 2026-06-25T05:53:02Z

Flaky tests detected in bab0c8a.
Some tests passed with failed attempts. The failures may not be related to this commit but are still reported for visibility. See the documentation for more information.

🔍 Workflow run URL: https://github.com/WordPress/gutenberg/actions/runs/28148665510
📝 Reported issues:

[Flaky Test] Pagination links are working as expected #78785 in /test/e2e/specs/editor/blocks/comments.spec.js
[Flaky Test] should update the HTML only for the latest navigation #78835 in /test/e2e/specs/interactivity/router-navigate.spec.ts

Absorb the inline-note tint test removal from 5b without applying it here: the per-author tinting feature (note-highlight-styles.js) lives on this branch, so the test passes and must stay. Recording the merge keeps future 5b->autosave cascades from re-deleting it.

adamsilverstein · 2026-06-30T22:18:15Z

Superseded by the fresh-stack restructure of Suggest mode (#73411).

The feature has been re-sliced from this 16-deep stack onto current trunk as a clean 2-PR stack:

Suggest mode: core subsystem (intent modes, overlay capture, structural + attribute suggestions) #79713 — Suggest mode: core (intent modes, overlay capture, structural + attribute suggestions, REST/PHP, auto-save, sidebar, summary)
Suggest mode: inline text and formatting suggestions via markers #79714 — Suggest mode: inline text suggestions (the <mark> marker layer)

The whole feature now sits behind a Suggestion Mode experiment (Settings → Experiments → Collaboration). Closing in favour of the new stack — details: #73411 (comment)

adamsilverstein and others added 30 commits March 23, 2026 13:11

Try: add support for multiple note ids

de57d01

Apply code improvements from Code Rabbit

6deb7d5

complete enabling multiple notes per block

0de44c6

ensure most recent selected when adding additional note

d3af188

remove blur guard

8e6e4e0

Revert "ensure most recent selected when adding additional note"

343333e

This reverts commit 7d2ff81.

Fix prettier formatting in comments.js

71a473a

Fix prettier formatting in collab-sidebar

7d19f06

Fix duplicate snackbar in multi-note E2E tests

4ca28f1

The first "Note added." snackbar persists when adding a second note, causing a strict mode violation (2 elements) and a premature metadata assertion. Dismissing the first snackbar before adding the second note fixes both tests.

Fix toolbar intercept in auto-select E2E test

11b8e1f

Merge branch 'trunk' into add/enable-multiple-notes-per-block

2890100

Resolve conflicts from PR #77614 (Notes refactor) by porting the multiple-notes-per-block feature into trunk's split component structure (notes.js, add-note.js, hooks.js, utils.js).

Merge branch 'trunk' into add/enable-multiple-notes-per-block

7548e9b

adamsilverstein added 2 commits June 18, 2026 10:58

Add Core backport changelog entry for inline note marker stripping

920a7be

Links to wordpress-develop#12215, which backports the render_block filter that strips inline note markers from rendered output.

adamsilverstein force-pushed the suggest-mode-autosave branch 2 times, most recently from cad24a9 to 41e40b1 Compare June 18, 2026 17:12

adamsilverstein added 2 commits June 18, 2026 19:57

Tests: keep elements block-support test version-agnostic after merge

45907a5

adamsilverstein force-pushed the phase-5b-collab-sidebar-actions branch from 9ec3f09 to 45907a5 Compare June 18, 2026 18:09

adamsilverstein added 3 commits June 18, 2026 20:10

adamsilverstein force-pushed the suggest-mode-autosave branch from 41e40b1 to d726b49 Compare June 18, 2026 18:11

adamsilverstein force-pushed the phase-5b-collab-sidebar-actions branch 2 times, most recently from e9d4d4f to 8d2858a Compare June 19, 2026 08:47

adamsilverstein added 8 commits June 20, 2026 08:46

Merge remote-tracking branch 'origin/phase-5b-collab-sidebar-actions'…

06b4cf2

… into suggest-mode-autosave

Merge remote-tracking branch 'origin/phase-5b-collab-sidebar-actions'…

227e9ba

… into suggest-mode-autosave # Conflicts: # phpunit/experimental/class-wp-rest-comments-controller-gutenberg-test.php # tools/eslint/suppressions.json

adamsilverstein changed the title ~~Suggest mode: Auto-save subsystem (Phase 6)~~ Suggest mode: Auto-save subsystem for pending suggestions Jun 30, 2026

adamsilverstein changed the title ~~Suggest mode: Auto-save subsystem for pending suggestions~~ Auto-save subsystem for pending suggestions Jun 30, 2026

adamsilverstein mentioned this pull request Jun 30, 2026

Suggest mode: core subsystem (intent modes, overlay capture, structural + attribute suggestions) #79713

Closed

adamsilverstein closed this Jun 30, 2026

Navigation Menu

Repository navigation

Auto-save subsystem for pending suggestions - #78308

Auto-save subsystem for pending suggestions#78308
adamsilverstein wants to merge 113 commits into
phase-5b-collab-sidebar-actionsfrom
suggest-mode-autosave

adamsilverstein commented May 14, 2026 •

edited

Loading

Uh oh!

adamsilverstein commented Jun 18, 2026

Uh oh!

github-actions Bot commented Jun 25, 2026

Uh oh!

adamsilverstein commented Jun 30, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

adamsilverstein commented May 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Overview

What's in this PR

Test plan

Stack

🗺️ PR Stack Navigation

Uh oh!

adamsilverstein commented Jun 18, 2026

Uh oh!

github-actions Bot commented Jun 25, 2026

Uh oh!

adamsilverstein commented Jun 30, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

adamsilverstein commented May 14, 2026 •

edited

Loading