Skip to content

fix(files): bound the file reads and name why one failed - #10834

Closed
rnoack1 wants to merge 1 commit into
kirodotdev:mainfrom
rnoack1:fix/bound-file-reads-and-name-causes
Closed

rnoack1 wants to merge 1 commit into
kirodotdev:mainfrom
rnoack1:fix/bound-file-reads-and-name-causes

Conversation

@rnoack1

@rnoack1 rnoack1 commented Sep 14, 2026 •

Copy link
Copy Markdown
Contributor

Problem / Motivation

Four file endpoints — file search, the project tree, and the two directory listings — were fetched with no client-side deadline. A wedged or slow gateway left the request unsettled, so the UI could not tell "still walking" from "no results", and every failure rendered as one generic error regardless of cause.

Their refusals also carried no machine-readable field: a 403 for a sensitive path and a 404 for a missing project root were distinguishable only by parsing untranslated server prose.

Why it matters

A caller that cannot tell a timeout from a denial cannot offer the right remedy — Retry is correct for a timeout and useless for a refusal, and rendering the server's English string in place of translated copy is a worse bug than a spinner.

This unit ships the bound, the classifier, and the first consuming surface: the folder panel's two failure notices. It is one of six units split out of a larger file-search timeout change that was too big to review in one piece.

The shared retry affordance (RetryControl) and the two keyboard hand-offs on useListKeyboardNav are not in this PR. An earlier revision carried them with no caller; they were deleted, and each lands with the surface that calls it.

What changed (motivation → approach → change)

  • withDeadline states the scope boundary deciding which fetches are bounded, and rejects with a TimeoutError-named Error so callers can key on the name.
  • The API client gains two deadline constants, each sized from a measurement recorded beside it: 15s for the tree walks (worst measured 4.20s on an 877k-dir tree — 28% of budget) and 10s for the listings (widest directory on the dev host, 8,974 entries, listed in 3.1ms — 0.03% of budget, local storage only).
  • The four backend responses gain a code field (access_denied, project_not_found).
  • A new searchErrorCause classifies a rejection as timed_out / failed / denied / root_missing, keyed on that code and never on the human string. A 403 carrying authRequired is a session expiry, so it is not reported as a refusal of the path. Two key maps consume it — LISTING_FAILURE_KEYS and SEARCH_FAILURE_KEYS — diverging on timed_out alone, because that is the only string naming which read expired; a refusal is the same fact whichever read hit it. CAUSE_BY_CODE carries only the two codes these endpoints actually emit, so no row is mapped without a rejection that reaches it.
  • FolderPanel renders both failure arms through those maps — the search arm via SEARCH_FAILURE_KEYS[searchErrorCause(searchError)], the listing arm via LISTING_FAILURE_KEYS[searchErrorCause(error)] — so a timeout, a refusal and a missing root each name themselves in the user's language instead of printing the gateway's English.
  • Both notices pass report={findReport(...)}. The "Ask the agent" hand-off resolves its structured context by exact-matching the displayed message against the error journal, whose key is the raw server sentence; translating the notice moves the text off that key, so without the explicit report the hand-off would degrade to a bare message and drop endpoint, status and code.
  • Four new notice strings in all 13 locale catalogs, which the i18n gate requires once the copy is translated rather than passed through.
  • A local capture harness (website/capture/listing-failure-copy.*, website/scripts/capture-listing-failure-copy.mjs) photographs and asserts all eight notices across en/ja and dark/light. Nothing in CI runs it; it is the repo's established review-evidence mechanism.

Every piece in this diff has a consumer inside this diff. The interim cost of deferring Retry is worth naming: until the surface that owns it lands, a timed-out listing or search offers "Ask the agent" as its only action, and the user's other recourse is to retype the query or reopen the panel.

Screenshots

Captured by this PR's own harness (node scripts/capture-listing-failure-copy.mjs) at the
revision under review. The 14 frames are not committed -- temp-screenshots/ is gitignored,
as the harness itself states -- and are served below from the capture revision
6bb55376e55b; five representative ones inline. The harness
asserts the translated copy is present and the raw deadline exceeded text is absent in every
scene, and exits non-zero otherwise.

Notice Frame
Listing timed out (en, dark) Listing timed out (en, dark)
No access to this folder, search arm (en, dark) No access to this folder, search arm (en, dark)
Folder not found (en, dark) Folder not found (en, dark)
Search timed out (ja, dark) Search timed out (ja, dark)
Folder not found (en, light) Folder not found (en, light)

The ja frame is the load-bearing one: both the notice copy and the "Ask the agent" control
are localised where the previous code printed the gateway's English. The light-theme frame
covers the contrast question the harness header raises.

Remaining frames at that same capture revision: the other five en dark states (both arms
x denied / root missing / failed), a second ja listing pair, and two further light-theme
scenes.

Tests

  • withDeadline.test.ts pins the rejection shape a replacement transport must satisfy — isDeadlineError, the cause mapping, and that a deadline rejection is not retried — so a future transport default cannot silently degrade every cause-keyed notice back to generic copy.
  • ApiClient.coverage.test.tsx asserts the fetch receives the deadline's signal rather than the caller's, that a caller passing no signal is bounded all the same, and that the tree read does not share the listings' shorter bound.
  • test_file_search.py asserts the UI-facing code rather than the human string.
  • error-code-baseline.json was regenerated with python test/test_error_code_contract.py --update; the counts move by the four responses this adds.

Manual verification

Verified locally at this revision: tsc -p tsconfig.app.json clean (break-armed by renaming the searchErrorCause export, which fails at FolderPanel.tsx), the four frontend suites below green, and the backend error-code contract green including test_baseline_is_not_stale. The CI lanes covering it are Frontend Lint & Type Check, Frontend Tests (1)–(4) and Frontend Coverage Merge. The frontend suites covering this change are withDeadline.test.ts, ApiClient.coverage.test.tsx, FolderPanel.search.test.tsx and FolderPanel.test.tsx; test_file_search.py covers the backend code field.

Pattern harvest

Rule candidate: semgrep

Pattern: err.message || t(key) — a truthy error message shadows the translated fallback, so
any rejection carrying server prose renders untranslated while the fallback key looks like it
covers the case. Both call sites this PR rewrites had exactly that shape. Flag || where the
left operand is an error .message and the right operand is a translation call.

Second candidate: a translated string passed where a lookup key is expected. ErrorNotice
resolves its agent hand-off by exact-matching message against the error journal, so
message={t(...)} without an explicit report= silently drops endpoint/status/code. Flag a
message prop whose value is a translation call on a component that also accepts report.

@github-actions github-actions Bot added fork Pull request from a fork (external contributor) readiness: action required A blocking check or review needs attention labels Sep 14, 2026
@dwu96

dwu96 commented Sep 14, 2026

Copy link
Copy Markdown
Contributor

👋 Hi! This PR is currently in draft status. Workflow runs won't be auto-approved until it's marked as ready for review.

When you're ready, click "Ready for review" and the workflows will be approved on the next cycle automatically.

@github-actions

github-actions Bot commented Sep 15, 2026 •

Copy link
Copy Markdown
Contributor

GPT 5.6 Review (fork) — ✅ no blocking findings

Reviewed 8fe347d81ff18dd717fb6c11108f8168348c9d8c via the fork AI-review pipeline; updated in place on each push.

Review details

No findings.
[GPT-REVIEWED] 8fe347d

@github-actions

github-actions Bot commented Sep 15, 2026 •

Copy link
Copy Markdown
Contributor

Design Review (Fable 5, fork) — 🔴 BLOCK (blocking)

Design-level review of 8fe347d81ff18dd717fb6c11108f8168348c9d8c via the fork AI-review pipeline — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

Design-Verdict: BLOCK

Cannot evaluate: missing screenshot of this revision's FolderPanel failure notices — every frame is hosted off capture commit 6bb55376, outside this PR.

Blockers

Rendered evidence is not admitted for this revision.
The diff reshapes a user-visible surface — a previously never-settling spinner becomes four cause-keyed notices in 13 locales, dark and light, with an "Ask the agent" report hand-off (FolderPanel.tsx, the new locale keys). The description serves its 14 frames from raw.githubusercontent.com/rnoack1/KiroCrew/6bb55376e55b.../temp-screenshots/... and itself says they are "served below from the capture revision 6bb55376e55b" with temp-screenshots/ gitignored — a commit outside this PR, not a user-attachments asset and not a committed image at head 8fe347d, so the evidence list names zero admitted artifacts. The shape questions the PR's own harness raises (JA copy fitting the 420px panel, light-theme notice contrast, hand-off control placement) are therefore unanswerable, and images on an unreviewable foreign commit could show any code.
Fix: commit the frames under temp-screenshots/ or .github/screenshots/ in this PR at head — the stated evidence path for a fork contributor whose --attach GitHub refuses — or re-host them as github.com/user-attachments assets.
Clears when: the evidence workflow's list names downloaded attachments or committed images for the eight notice states at the PR head SHA.

Watch

The 10s listing bound is sized from local storage only (the constant's comment concedes "no network mount was available to sample"), and Retry is deferred to a later unit — so a slow-but-working network-mounted folder degrades from an eventual listing to a hard "timed out" notice whose only action is the agent hand-off.
Clears when: a network-mount measurement validates BROWSE_FILES_TIMEOUT_MS, or the RetryControl unit lands so a timeout has an in-panel remedy.

[DESIGN-REVIEWED] 8fe347d

@github-actions

github-actions Bot commented Sep 15, 2026 •

Copy link
Copy Markdown
Contributor

First Principles Review (Fable 5, fork) — ✅ PASS

Premise-level review of 8fe347d81ff18dd717fb6c11108f8168348c9d8c via the fork AI-review pipeline — why this exists and whether the shipped surface is the smallest honest version. Updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

First-Principles-Verdict: PASS

Verify the 10s listing bound on a network-mounted directory — the one case the sizing measurement itself admits it never sampled.

Not justified as shipped

  • Item 8 — undeclared: every other caller of these endpoints (ProjectPicker, WorkspacePicker, FilePickerMenu, FileBrowserRail — 4 non-test surfaces grepped for api.browse*/api.fileSearch/api.projectTree) now fails at the deadline instead of spinning; the description names only the folder panel. Verified harm-free: each renders its own translated generic copy, so no raw deadline exceeded leaks there.

What this change ships

Inventory (9 items) — 8 justified

Intent: FIX — bound the four file-read fetches and name each failure's cause in translated copy, folder panel first.

  1. File search, project tree and both listings give up at 15s/10s instead of spinning forever — justified
  2. Folder-panel failure notices name the cause in the user's language, not the server's English — justified
  3. "Ask the agent" from those notices still carries endpoint/status/code — justified
  4. An expired-session 403 reads as generic failure, not a refused folder — justified
  5. The four backend refusals carry a machine code (ratchet worklist moves −4) — justified
  6. Shared searchErrorCause classifier + two key maps (no prior classifier exists; grep errorCause|FAILURE_KEYS = 0) — justified
  7. Four notice strings in all 13 catalogs (i18n gate) — justified
  8. Other pickers/rail inherit the bound and end their spinners at the deadline — undeclared
  9. Local capture harness asserting all eight notices (648 sibling capture scripts; declared) — justified

[FIRST-PRINCIPLES-REVIEWED] 8fe347d

@github-actions

github-actions Bot commented Sep 15, 2026 •

Copy link
Copy Markdown
Contributor

UX Review (Fable 5, fork) — 🔴 BLOCK

UX-level review of 8fe347d81ff18dd717fb6c11108f8168348c9d8c via the fork AI-review pipeline — updated in place on each push. A BLOCK verdict blocks PR readiness; PASS/CONCERNS are advisory.

UX-Verdict: BLOCK

Cannot evaluate: missing admitted screenshots — every frame is a raw fork-commit URL pinned outside this PR, which the evidence step does not admit.

Evidence gaps

  • Listing-arm notices ("Folder listing timed out", "No access to this folder", "Folder not found", "Unable to list folder" fallback) — no admitted image; the raw.githubusercontent.com/.../6bb55376.../temp-screenshots/... links failed to close the gap (commit outside this PR, not a user-attachments asset).
  • Search-arm notices ("Search timed out", "No access to this folder", "Folder not found", "Search failed") — same artifact, same failure.
  • Localised state (ja) and light-theme contrast state the PR itself calls load-bearing — same artifact, same failure.
  • Close by committing the harness frames under temp-screenshots/ on this PR's branch (git add -f; the workflow reads committed media for fork PRs) or attaching them as user-attachments assets.

Watch

  • A timed-out listing/search offers only "Ask the agent" — retry means retyping the query or reopening the panel. The PR names this interim cost and defers RetryControl deliberately; a human should confirm that sequencing is acceptable to ship.

[UX-REVIEWED] 8fe347d

@github-actions

github-actions Bot commented Sep 15, 2026 •

Copy link
Copy Markdown
Contributor

Opus 5 Review (fork) — ✅ no blocking findings

Reviewed 8fe347d81ff18dd717fb6c11108f8168348c9d8c via the fork AI-review pipeline; updated in place on each push.

Review details

api.projectTree is bounded at 15s by a rationale that cites walk ceilings the endpoint does not apply — no blocking issue.

FINDING — website/src/api/client.ts:4126 — projectTree: (path: string) => withDeadline(FILE_SEARCH_TIMEOUT_MS, …) is sized by a comment claiming the walk is ceilinged at _WALK_MAX_DIRS_VISITED/_WALK_MAX_SCAN_SCOPED "so no tree size can" exhaust 15s, but those ceilings are read only by the file-search walk (handlers/files.py:4637, :4668); api_project_tree spends up to timeout=5 (rev-parse) + timeout=15 (ls-files, files.py:8306) before its fallback runs two uncapped os.walks, so on any project that reaches that fallback the client aborts at 15s before the server starts the walk → useTreeState returns 'error', the rail shows "Couldn't load the file tree" and FolderPanel silently drops to the flat one-level listing where the tree previously rendered → Fix: give projectTree its own constant covering the endpoint's real budget (the 20s of git timeouts plus the walk), or drop the withDeadline wrap on it — this PR's copy and cause keys cover the search and listing arms only, not the tree read.

FINDING — website/capture/listing-failure-copy.tsx:93 — url.startsWith('/api/project-tree') and '/api/project-git-status' match no URL the client builds (/api/project/tree, /api/project/git/status), so with path === projectDir === PROJECT making atProjectRoot true, useTreeState's read falls through to realFetch and out to the Vite /api proxy on every scene; the 16 assertions then pass only because that escaped request errors, since a matching 200 {paths: []} makes treeState 'ready' and treeMode replaces both the listing and search arms the frames exist to photograph → Fix: match the real prefixes and answer the tree read with a refusal (e.g. json(403, …)), not a 200.

[OPUS-REVIEWED] 8fe347d

@github-actions github-actions Bot added the merge conflict Branch has merge conflicts with its base — author must resolve before merge label Sep 15, 2026
@rnoack1
rnoack1 force-pushed the fix/bound-file-reads-and-name-causes branch from 130328f to 9207d18 Compare September 15, 2026 23:45
@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running merge conflict Branch has merge conflicts with its base — author must resolve before merge labels Sep 15, 2026
@rnoack1
rnoack1 force-pushed the fix/bound-file-reads-and-name-causes branch from 9207d18 to d2edb79 Compare September 16, 2026 01:03
@github-actions github-actions Bot added readiness: action required A blocking check or review needs attention and removed readiness: action required A blocking check or review needs attention labels Sep 16, 2026
@rnoack1
rnoack1 force-pushed the fix/bound-file-reads-and-name-causes branch from d2edb79 to 41bb7ca Compare September 16, 2026 01:47
@github-actions github-actions Bot added readiness: checking Automated validation is still running and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running labels Sep 16, 2026
@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention and removed merge conflict Branch has merge conflicts with its base — author must resolve before merge readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running labels Sep 19, 2026
@rnoack1
rnoack1 force-pushed the fix/bound-file-reads-and-name-causes branch from 0b5b1af to 66aab18 Compare September 22, 2026 05:08
@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running labels Sep 22, 2026
@rnoack1
rnoack1 force-pushed the fix/bound-file-reads-and-name-causes branch from 66aab18 to b22501d Compare September 22, 2026 06:08
@github-actions github-actions Bot added readiness: checking Automated validation is still running readiness: action required A blocking check or review needs attention merge conflict Branch has merge conflicts with its base — author must resolve before merge and removed readiness: action required A blocking check or review needs attention readiness: checking Automated validation is still running labels Sep 22, 2026
@rnoack1
rnoack1 force-pushed the fix/bound-file-reads-and-name-causes branch from b22501d to e04e324 Compare September 22, 2026 20:53
@github-actions github-actions Bot added merge conflict Branch has merge conflicts with its base — author must resolve before merge and removed merge conflict Branch has merge conflicts with its base — author must resolve before merge labels Sep 22, 2026
@rnoack1
rnoack1 force-pushed the fix/bound-file-reads-and-name-causes branch from e04e324 to 22b5c95 Compare September 23, 2026 05:55
@github-actions github-actions Bot added readiness: checking Automated validation is still running and removed readiness: action required A blocking check or review needs attention merge conflict Branch has merge conflicts with its base — author must resolve before merge readiness: checking Automated validation is still running labels Sep 23, 2026
@bolichen97

Copy link
Copy Markdown
Contributor

Consolidation: this PR and PR #7068 carry the same change

This PR and PR #7068 (same author, still being driven) ship the same fix, mostly byte for byte. Merging the two heads with git merge-tree gives conflicts in 9 files, including an add/add on website/src/lib/searchErrorCause.ts, so only one of them can land as written.

What is shared (line numbers at 22b5c95):

What #7068 has beyond this PR: its withJournaledDeadline (client.ts:2145 there) journals timeouts with code: 'timeout' and the endpoint. Here a timed-out read is not journaled, so findReport('deadline exceeded') returns nothing and "Ask the agent" loses the endpoint. #7068 also adds not_a_directory on both listing 400s, the extra CAUSE_BY_CODE rows and the "Refresh to retry" remedy.

What only this PR has:

  1. FolderPanel.search.test.tsx:362, the describe block "hands the agent the structured report, not the translated copy". Its two tests assert that the ERROR_HANDOFF_KEY prompt carries HTTP 403 and access_denied for the search and listing arms. fix(chat): bound the @-menu file-search fetch so a slow gateway fails visibly #7068 pins the hand-off only for the tree notice (FolderPanel.tree.test.tsx:247).
  2. FolderPanel.search.test.tsx:341: the session-expiry test builds its error with a coded body, JSON.stringify({ code: 'access_denied' }), and authRequired=true. fix(chat): bound the @-menu file-search fetch so a slow gateway fails visibly #7068's version (FolderPanel.search.test.tsx:362 at df258de) uses { error: 'Access denied' }, so it still passes if || err.authRequired is deleted from searchErrorCause. Adding the code there is a one-line fix.
  3. Optional: the capture harness, and the two semgrep candidates under "Pattern harvest". fix(chat): bound the @-menu file-search fetch so a slow gateway fails visibly #7068's body calls its own pattern "Not generalizable".

Suggestion: host in #7068. It was opened earlier, it is fresher (199 commits behind origin/main vs 289 here), and it is a superset. Port items 1 and 2 into it; bring item 3 over only if reviewers want it. Leave out browseDirs(path, signal?): #7068 dropped it because no production caller passes a signal. Add it back only if PR #10844 lands on top, because its tests expect an AbortSignal there. PR #10838, PR #10839 and PR #10844 import from this PR's lib/searchErrorCause. #7068 exports every symbol they use, so they can import from #7068 instead.

From a full open-PR audit against origin/main c66f95b (2026-09-25); findings were checked by independent reviewers reading the code. Reply here if this is wrong.

Adds a client-side deadline to the file-search and directory-listing reads, and a machine-readable code on their refusals, so a caller can tell a timeout from a denial instead of rendering one generic error.
@NicholasRBowers

Copy link
Copy Markdown
Contributor

🤖 Kiro Crew [operator: NicholasRBowers#a942f9ca]: This PR has been inactive for 7+ days. I reviewed the blockers but they require your input:

  • Most of this change has already landed via your own PR fix(chat): bound the @-menu file-search fetch so a slow gateway fails visibly #7068, which merged on 2026-09-29 (merge commit 18243811). The consolidation audit on this PR recommended hosting the work in fix(chat): bound the @-menu file-search fetch so a slow gateway fails visibly #7068 as the superset, and that is what happened: website/src/lib/searchErrorCause.ts and website/src/lib/withDeadline.ts are both on main now, so this branch's versions are add/add conflicts against a merged superset rather than new files. Driving this head to green would mean resolving nine files of conflict where the content is already upstream, so the decision of what this PR is still for has to come first.
  • What the audit found unique to this PR, and still worth porting, if you want it: (1) the FolderPanel.search.test.tsx describe block "hands the agent the structured report, not the translated copy", whose two tests assert the ERROR_HANDOFF_KEY prompt carries HTTP 403 and access_denied for the search and listing arms — fix(chat): bound the @-menu file-search fetch so a slow gateway fails visibly #7068 pins the hand-off only for the tree notice; (2) the session-expiry test building its error with a coded body (JSON.stringify({ code: 'access_denied' })) plus authRequired=true, which fails if || err.authRequired is deleted from searchErrorCause, where fix(chat): bound the @-menu file-search fetch so a slow gateway fails visibly #7068's version would still pass. Optionally (3) the capture harness.
  • So the call is yours: close this PR, or reduce it to those residual test deltas on top of current main. Either is a scope decision, not a defect I can patch, so I'm not touching the branch.
  • For the record, if it does continue: Design Review and UX Review both BLOCK because every evidence frame is served from raw.githubusercontent.com/rnoack1/KiroCrew/6bb55376…/temp-screenshots/…, a commit outside this PR, so no artifact is admitted. UX Review names the remedy for a fork PR — commit the harness frames under temp-screenshots/ on this branch with git add -f, since the workflow reads committed media for forks. Opus also has one non-blocking FINDING: projectTree is bounded at 15s by a comment citing walk ceilings the endpoint does not apply, so a project reaching the uncapped os.walk fallback aborts client-side before the server starts walking.

When you've addressed these, the pipeline will re-assess on its next cycle.

Copy link
Copy Markdown
Contributor

PR triage: SUPERSEDED by #7068 (merged 2026-09-29, 18243811d6). Closing.

I checked this PR's diff (merge base 20ed6a1dc7, head 8fe347d81f) against origin/main at 5edf2feecb. Every piece this PR ships is already on main, either the same or in a fuller form:

This PR On origin/main
files.py 403 code: "access_denied" on file search and both listings, 404 code: "project_not_found" dashboard/file_api/search.py:212 (access_denied), :224 (project_not_found); handlers/files.py:2841, :2895 (access_denied), plus not_a_directory
test_file_search.py asserts on data["code"] test/test_file_search.py:188, :216
FILE_SEARCH_TIMEOUT_MS 15s / BROWSE_FILES_TIMEOUT_MS 10s wrapping the file reads website/src/api/client/files.ts:46, :68, now through withJournaledDeadline, so a timed-out read gets journaled and "Ask the agent" keeps the endpoint
lib/searchErrorCause.ts (searchErrorCause, LISTING_FAILURE_KEYS, the authRequired gate) website/src/lib/searchErrorCause.ts:19-100, plus TREE_FAILURE_KEYS, failureMessage, RETRYABLE_SEARCH_CAUSES
withDeadline TimeoutError shape and its tests website/src/lib/withDeadline.ts:30; withDeadline.test.ts:31, :38
FolderPanel search and listing arms rendered from the cause maps FolderPanel.tsx:37, :577, :674
Residual 1 from the consolidation audit (the "hands the agent the structured report" tests) FolderPanel.search.test.tsx:672 onward: both the search and listing cases assert HTTP 403 and access_denied
Residual 2 (session-expiry test with a coded body) FolderPanel.search.test.tsx:359-367: body carries code: 'access_denied' with authRequired=true

Both residual test deltas from the consolidation comment have already been ported. The only parts not on main are the local capture harness (website/capture/listing-failure-copy.*, scripts/capture-listing-failure-copy.mjs) and the semgrep ideas in the description. Those are review-evidence tooling, not part of the PR's goal. If you want them, a small new PR against current main is the cleanest route.

git cherry shows the head is not patch-equal to main, because #7068 was a different squash. Rebasing would mean resolving 9 add/add and content conflicts in favour of code that is already upstream. No breaking change and no P0 bug found.

Reopen if I've missed something this PR still carries.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

fork Pull request from a fork (external contributor) merge conflict Branch has merge conflicts with its base — author must resolve before merge needs-author-decision PR blocked on author input

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants