• closeBtnLabel (driver and step level) to set the close button's aria-label, for translated tours. Defaults to Close. (#623)
• overlayClickBehavior: "none" to ignore overlay clicks while Escape and the close button still close the tour. (#435)

• Keyboard control follows the popover buttons: arrow keys do nothing when their button is hidden or disabled, and Escape does nothing when the close button is disabled. (#589)
• The current popover is hidden while waitForElement waits for the next step's element, so a stale step no longer stays on screen. (#578)
• Only one instance is on screen at a time. Starting a tour or highlight closes any other active instance; it gets onDeselected and onDestroyed but not onDestroyStarted.

• destroy() on an instance that is not active is now a no-op instead of tearing down another instance's tour. (#504)
• The highlighted element's own aria-controls, aria-expanded and aria-haspopup are restored when the highlight moves on or the tour ends, instead of being removed. (#495)
• No more small white square between steps without an element. (#549)
• onDeselected now fires when moving between two steps on the same element, or between two steps without an element.
• The driver-level disableButtons now applies to tour steps. It used to be ignored during a tour. A step's own disableButtons still replaces it.

• closeBtnLabel (driver and step level) to set the close button's aria-label, for translated tours. Defaults to Close. (#623)
• overlayClickBehavior: "none" to ignore overlay clicks while Escape and the close button still close the tour. (#435)

• Keyboard control follows the popover buttons: arrow keys do nothing when their button is hidden or disabled, and Escape does nothing when the close button is disabled. (#589)
• The current popover is hidden while waitForElement waits for the next step's element, so a stale step no longer stays on screen. (#578)
• Only one instance is on screen at a time. Starting a tour or highlight closes any other active instance; it gets onDeselected and onDestroyed but not onDestroyStarted.

• destroy() on an instance that is not active is now a no-op instead of tearing down another instance's tour. (#504)
• The highlighted element's own aria-controls, aria-expanded and aria-haspopup are restored when the highlight moves on or the tour ends, instead of being removed. (#495)
• No more small white square between steps without an element. (#549)
• onDeselected now fires when moving between two steps on the same element, or between two steps without an element.
• The driver-level disableButtons now applies to tour steps. It used to be ignored during a tour. A step's own disableButtons still replaces it.

• Add the filter option to the import*() methods of ZipFS and ZipDirectoryEntry: the function receives each entry read from the zip file and returns true (or a promise resolving to true) to import it. It can read the data of the entry to decide, and the entries left out never reach the duplicates policy
• Add the filter option to the export*() methods and to getExportedSize(): the function receives each ZipEntry of the tree and returns true to export it. Leaving out a directory leaves out its subtree, the entries left out are not read and are not counted by onprogress and onentryprogress
• Add the filter option to exportFileSystemHandle(), with the same semantics as the zip exports
• Add the filter option to addFileSystemHandle() and addFileSystemEntry(): the function receives each handle found and its path, so a directory like node_modules or hidden files can be left out without walking them

Full Changelog: https://github.com/gildas-lormeau/zip.js/compare/v2.22.1...v2.23.0

Co-Authored-By: Claude Fable 5.1 noreply@anthropic.com

• A zip file written behind a prefix which was removed together with its whole first entry, like a self-extracting page whose first entry is the face shown by the host, is read again: the entries are shifted to their real positions and their data can be read. Since v2.21.0, the shift was decided on the first central directory record only, and a first record whose local file header lies in the removed bytes proves nothing, so the entries kept their stored offsets and getData() failed with ERR_LOCAL_FILE_HEADER_NOT_FOUND, whatever the strictness option. The reader now checks the following records until one settles the question, and keeps the protection against a damaged end of central directory record whose local file headers are right
• An entry which lies before the start of the zip file after such a shift is no longer reported as WARNING_UNSORTED_CENTRAL_DIRECTORY

Full Changelog: https://github.com/gildas-lormeau/zip.js/compare/v2.22.0...v2.22.1

Co-Authored-By: Claude Fable 5.1 noreply@anthropic.com

• Optimize esm exports by shifting the protocol burden to getters: #98932
• fix: fix pnpm builds in Docker examples: #97352
• Assert transient task references are not serialized: #99300
• Reorganize bundle analyzer toolbar: #98920
• Show async module scope in bundle analyzer: #98918
• Show bundle changes in compare treemap: #98837
• improve disk read semantics around MustExist and AllowMissing: #99249
• Allow opt-in facade-backed export mangling: #99309
• Add bundle analyzer route summary: #98787
• [turbo-tasks-backend] Correctly maintain outdated_collectibles to prevent unnecessary stale executions: #99386
• Upgrade React from 8b0da1c6-20260922 to 278794d7-20261002: #99600
• Warn when Partial Prefetching is not configured: #99452
• Store the Turbopack module cache in a Map: #98947
• Improve agent feedback reporting reliability: #99591
• Enable empty generateStaticParams deploy diagnostics: #99406
• test: enable Nx deployment coverage: #99407
• test: enable deployment coverage for monorepo fixtures: #99395
• test: capture deployment experimental flags in config: #99534
• test: extract a shared helper for deployment feature flags: #99533
• Move App Page header parsing to handler entrypoints: #99506
• docs: clarify ensureStatic levels from dogfooding: #99590
• Gate param-matching deploy tests on the adapter: #99582
• docs: update the MCP guide for next-devtools-mcp 0.4: #99583
• Document agent upgrades across Next.js: #99293
• [test] Fix the legacy image preload order assertion for React 18: #99572
• docs: Remove task decomposition and validation boilerplate from AGENTS.md: #99558
• Turbopack: Use a FrozenMap for DiskFileSystemMap: #99569
• [test] Expose first-run e2e test failures in CI: #99528
• Turbopack: resolveFileUrl/placeholderFileUrl is always called with a module path: #99560
• Turbopack: Remove implicit mismatched root fallthrough for AfterResolvePluginCondition, accept None as a root: #99556
• Turbopack: Allow ResolveMap entries to match against all possible roots: #99555
• Turbopack: Support import.meta.url in additional roots: #99511

Huge thanks to @lukesandberg, @6iu8a, @wbinnssmith, @sokra, @aurorascharff, @jamiboym, @gnoff, @devjiwonchoi, @unstubbable, and @bgw for helping!

• 4b1c8d3ea2b3de1a2ae852fd5631e33d28a82b15 #4721 Thanks @giaBaoJS! - fix: assigning length on an observable array now respects intercept. A cancelled or modified splice used to be overridden by a direct write to the backing array, which truncated it without notifying observers and made the next mutation throw "Modification exception". A non-integer length now throws "Out of range" before touching the array.

• 60be47cab0926d12f22449ec6c809fb5ec6601c4 #4716 Thanks @AcEKaycgR! - fix: observable.box(array) no longer warns under observableRequiresReaction when a spy is active. The create spy event now reports newValue as the value itself rather than a string, matching the declared IBoxDidChange type. Previously the constructor stringified the value, and stringifying an observable array reads its length, which counted as an observable read outside a reactive context. Spy listeners that relied on newValue being a string for create events will now receive the raw value.

• 3551aeb051ff922717b03c4fbc04a6b98f7f3b3e #4717 Thanks @breken-ai! - Remove the signal abort listener when a reaction disposes itself (r.dispose() or a resolved when), so long-lived AbortSignals no longer accumulate listeners that keep disposed reactions in memory

• 7e4e27dbfbf0b56770a11e9cba913e5d49c0c6f8 #4718 Thanks @breken-ai! - observe/spy add events of an observable Set now report the value that was actually stored (e.g. the observable copy of a plain object), matching ObservableMap and observable arrays, so set.has(change.newValue) holds

2026-10-02

• Improve try-catch deoptimization to cover calling methods on objects (#6541)

• Fix a situation where some feature-detections of core-js were not triggering properly (#6541)

• #6541: Retain callbacks of helper methods called from try statements (@lukastaegert)
• #6542: Run additional tests as part of the coverage job (@lukastaegert)
• #6544: Update msys2/setup-msys2 digest to ec48f7c (@renovate[bot], @lukastaegert)
• #6545: Update minor/patch updates (@renovate[bot])
• #6546: Update dependency @mermaid-js/mermaid-cli to v12 (@renovate[bot])
• #6547: Update Rust crate swc_compiler_base to v65 (@renovate[bot])

• Updated dependencies []:
  • rrweb-player@2.1.7

• Updated dependencies []:
  • rrdom@2.1.7
  • @rrweb/types@2.1.7