{ "commits": { "7862541": { "author": "Samba Murthy Bandaru", "avatar": "https://lh5.googleusercontent.com/-IXL8leFFF50/AAAAAAAAAAI/AAAAAAAAAAA/gTMLUUxtqLE/s120-p/photo.jpg", "description": "Add visual_caret_movement.{h,cc} implementing the core algorithm that\nwalks inline fragments to find the visually adjacent caret position,\ntracking bidi embedding levels across keystrokes to disambiguate caret\nplacement at bidi boundaries.\n\nKey components:\n- VisualFragment: lightweight wrapper around InlineCursor for visual\n fragment traversal within a single line.\n- FindFragmentForPosition: bypasses ComputeInlineCaretPosition to avoid\n BidiAdjustment that corrupts fragment identification at bidi boundaries.\n- MoveCaretVisuallyRight/Left: three-step algorithm (resolve fragment,\n advance within fragment, cross to next visual fragment).\n- Grapheme-cluster-aware offset advancement.\n- Cross-line movement at line boundaries.\n\nWire the visual movement into SelectionModifier behind the\nBidiVisualOrderCaretMovement runtime feature flag (disabled by default,\ndepends on BidiCaretAffinity).\n\nAdd comprehensive test suite exercising: pure LTR/RTL traversal, mixed\nbidi boundary crossing, multiple bidi runs, Unicode bidi controls,\nsurrogate pairs, combining characters, ZWJ emoji, CSS bidi-override,\natomic inlines, cross-line movement, and anti-loop guarantees.\n\nExplainer: https://github.com/MicrosoftEdge/MSEdgeExplainers/blob/main/Editing/bidi-visual-order-caret-movement-explainer.md\nDesign-Doc: https://docs.google.com/document/d/16WdeXHgbiQFXxiu1CfCGU4AEMBN9y3ssAQSYjw3qaJc/edit?usp=sharing\nIntent-to-Prototype: https://chromestatus.com/feature/5074816756482048\nBug: 499819853\n\nChange-Id: I55211ace39f31be4690a98a419a84fedbad2dc8a\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7862541\nCommit-Queue: Sambamurthy Bandaru \u003csambamurthy.bandaru@microsoft.com\u003e\nReviewed-by: Shweta Bindal \u003cshwetabindal@microsoft.com\u003e\nReviewed-by: Kent Tamura \u003ctkent@chromium.org\u003e\nCr-Commit-Position: refs/heads/main@{#1646508}", "email": "sambamurthy.bandaru@microsoft.com", "hash": "64eb1c04d6e2c767af754d0eca4e34128442ac2c" }, "7862620": { "author": "Samba Murthy Bandaru", "avatar": "https://lh5.googleusercontent.com/-IXL8leFFF50/AAAAAAAAAAI/AAAAAAAAAAA/gTMLUUxtqLE/s120-p/photo.jpg", "description": "In FrameSelection::Modify(), persist caret_bidi_level_ and\nentered_bidi_run_ across keystrokes and override selection affinity\nbased on bidi level transitions to ensure correct caret rendering\nvia BidiCaretAffinity.\n\nAt bidi boundary offsets (where two fragments share an offset), affinity\ndetermines which fragment the caret resolves into:\n- Level transition: use Downstream (start of new run's fragment)\n- Same level, LTR: use Upstream (end edge of LTR fragment)\n- Same level, RTL: use Downstream (start edge of RTL fragment)\n\nReset bidi state on non-keyboard selection changes (mouse clicks,\nprogrammatic) in SetSelectionDeprecated() to prevent stale bidi level\nfrom corrupting subsequent visual movements.\n\nBug: 499819853\n\nChange-Id: I3b23a3b073e3d87351a7568d72a8a2c70c1e8892\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7862620\nReviewed-by: Kent Tamura \u003ctkent@chromium.org\u003e\nReviewed-by: Shweta Bindal \u003cshwetabindal@microsoft.com\u003e\nCommit-Queue: Sambamurthy Bandaru \u003csambamurthy.bandaru@microsoft.com\u003e\nCr-Commit-Position: refs/heads/main@{#1646521}", "email": "sambamurthy.bandaru@microsoft.com", "hash": "8bf9c73f18027f76627a52435c825aa616bf1255" }, "7899676": { "author": "tannal", "avatar": "https://lh4.googleusercontent.com/-YsWr1H8zLkk/AAAAAAAAAAI/AAAAAAAAAAA/Ph-EdhUt5EE/s120-p/photo.jpg", "description": "Split common properties for link elements \u003ca\u003e, \u003csvg\u003e and \u003carea\u003e into a mixin. This is in preparation for aligning \u003carea\u003e with the other link elements.\n\nAdd hreflang and type attributes to html area element behind a flag, currently enabled for test.\n\nWhen the flag is enabled both HTMLAnchorElement and HTMLAreaElement share the same HyperlinkElementUtils IDL mixin.\n\nSee https://github.com/whatwg/html/pull/12446\n\nI2S: https://groups.google.com/a/chromium.org/g/blink-dev/c/FEeoLfm2yg4\n\nChange-Id: I73f8b6ca87a3bd64217314d67a14c8eb376a0a82\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7899676\nReviewed-by: Frédéric Wang Nélar \u003cfwang@igalia.com\u003e\nCommit-Queue: Frédéric Wang Nélar \u003cfwang@igalia.com\u003e\nReviewed-by: Stephen Chenney \u003cschenney@chromium.org\u003e\nCr-Commit-Position: refs/heads/main@{#1646464}", "email": "mtan@igalia.com", "hash": "2693c880ae12b9e5b61887b348ae22924d5a769c" }, "7922184": { "author": "Hamza Mohammed", "avatar": "https://lh6.googleusercontent.com/--PVoUk_4gf4/AAAAAAAAAAI/AAAAAAAAAAA/ks5dVlsM1zk/s120-p/photo.jpg", "description": "This flag will allow us to test certain Gemini params such as different response ready intervals. It is a change that will help us decouple response ready intervals from the Gemini Copresence Enabled flag which will help in cleaning it up from the codebases (both Chromium and ios_internal).\n\nNote that this flag will NOT be visible to the user.\n\nBug: 522712050\nChange-Id: I3f3249d3729d7747ef4a27aeeb3e4fafbbc9caca\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7922184\nCommit-Queue: Hamza Mohammed \u003chamzamoh@google.com\u003e\nReviewed-by: Adam Arcaro \u003cadamta@google.com\u003e\nReviewed-by: Joe Peplowski \u003cjoepeplowski@google.com\u003e\nCr-Commit-Position: refs/heads/main@{#1646460}", "email": "hamzamoh@google.com", "hash": "a099554d225e3b85238133f9f616e3bd6c3459ae" }, "7922377": { "author": "Bofeng Chen", "avatar": "https://lh3.googleusercontent.com/-bOweUmWd05k/AAAAAAAAAAI/AAAAAAAAAAA/JswZwavqFpU/s120-p/photo.jpg", "description": "This is a reland of commit d3caa3bce2fffbfc6eb46b5ac7315b57abc27f30\nThe original CL was reverted because it was suspected to be the culprit CL for these two interactive ui tests:\nHelpBubbleHandlerInteractiveUiTest.ElementBecomesHiddenOnSecondaryUISwap and WallpaperSearchOptimizationGuideInteractiveTest.CustomizeButtonsWorkTogether failures.\n\nBut after the CL was reverted, at commit position #1646144, these two cases were still failing at #1646147. https://ci.chromium.org/ui/p/chromium/builders/ci/win11-arm64-rel-tests/7376/blamelist\n\nTherefore, the original CL was not the culprit for the test failure.\n\nOriginal change's description:\n\u003e [Bookmarks] Show a promo when the bookmark bar is auto-hidden.\n\u003e\n\u003e When the bookmark bar is automatically hidden due to inactivity, trigger\n\u003e the `kIPHBookmarkBarSimplifiedFeature` promo. Updates a unit test to\n\u003e verify this behavior.\n\u003e\n\u003e Make sure the promo shows and only once, and clicking the undo button\n\u003e restores the bookmark bar on NTP:\n\u003e https://screencast.googleplex.com/cast/NjY3NzQxNTcxMDQyNTA4OHw5MTg1MGZiZC0yOA\n\u003e\n\u003e If the toast is dismissed, the bookmark bar visibility state stays\n\u003e \"Always Hide\":\n\u003e https://screencast.googleplex.com/cast/NDU0OTcxOTI0NTE5MzIxNnwyNmE4ODYzMi1kOA\n\u003e\n\u003e Fixed: b:520023694\n\u003e Change-Id: If8312ec376f458e1584ba2d6071ae866699e8d78\n\u003e Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7922377\n\u003e Commit-Queue: Bofeng Chen \u003cbofengchen@google.com\u003e\n\u003e Auto-Submit: Bofeng Chen \u003cbofengchen@google.com\u003e\n\u003e Reviewed-by: Riley Tatum \u003crtatum@google.com\u003e\n\u003e Reviewed-by: Tibor Goldschwendt \u003ctiborg@chromium.org\u003e\n\u003e Cr-Commit-Position: refs/heads/main@{#1645751}\n\nChange-Id: Ibdb130071afe441862a15f3a18e38418d58d3dd6\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7934748\nReviewed-by: Tibor Goldschwendt \u003ctiborg@chromium.org\u003e\nCommit-Queue: Bofeng Chen \u003cbofengchen@google.com\u003e\nReviewed-by: Riley Tatum \u003crtatum@google.com\u003e\nCr-Commit-Position: refs/heads/main@{#1646515}", "email": "bofengchen@google.com", "hash": "e656d2d5179f1e10d02e19718eda2127e13eabb8" }, "7924846": { "author": "Dan Polanco", "avatar": "https://lh5.googleusercontent.com/-KFFbQTPR47s/AAAAAAAAAAI/AAAAAAAAAAA/POkoSj10t14/s120-p/photo.jpg", "description": "When creating tabs from the NTP and then navigating or swiping to a\nbrowsing layout, the composited bottom bar screenshot would remain\nstale.\n\nNative pages are permanently locked to SHOWN, which caused\nScrollingBottomViewResourceFrameLayout to defer all captures. Because\nthe controls will never be released, the captures never happen.\n\nThis CL resolves this by introducing BottomBarConstraintsSupplier, which\nobserves the active tab and overrides the constraints to BOTH when on\nthe NTP. This ensures ScrollingBottomViewResourceFrameLayout allows\nscreenshot updates, preventing stale screenshots.\n\nBug: 519148302\nChange-Id: I303638012b446e4d0da1a9898f97b00184c614f4\nFixed: 521261060, 519948927\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7924846\nCommit-Queue: Dan Polanco \u003cpolardz@google.com\u003e\nReviewed-by: Sky Malice \u003cskym@chromium.org\u003e\nCr-Commit-Position: refs/heads/main@{#1646523}", "email": "polardz@google.com", "hash": "b833ceb7b9921a34b7d641fa44d242f58e448e57" }, "7928088": { "author": "Jihad Hanna", "avatar": "https://lh4.googleusercontent.com/-v4uH-kSLLc0/AAAAAAAAAAI/AAAAAAAAAAA/Z4UwQ-pCWss/s120-p/photo.jpg", "description": "This CL refactors the form-filling logic to remove the\n`_pendingAutocompleteFieldID` state from `AutofillAgent`.\n\nInstead of caching the initiating field ID in C++ and passing it to\nJavaScript to force-fill it, we now determine the initiating field on\nthe JS side by checking `element === document.activeElement`.\n\nThis aligns the iOS implementation with Blink's behavior (which checks\n`element.Focused()`) and fixes a potential bug in multi-frame scenarios\nwhere local field IDs could collide across different frames.\n\nBug: 40266699\nChange-Id: I4502b03d055b44844ff933a8f45ddca1c0346c0d\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7928088\nReviewed-by: Vincent Boisselle \u003cvincb@google.com\u003e\nCommit-Queue: Jihad Hanna \u003cjihadghanna@google.com\u003e\nCr-Commit-Position: refs/heads/main@{#1646481}", "email": "jihadghanna@google.com", "hash": "d0bcccb51b8d779f99b70b3c01d4ea76bf080a60" }, "7929864": { "author": "PKI Metadata Updates Bot", "avatar": "https://lh3.googleusercontent.com/-XdUIqdMkCWA/AAAAAAAAAAI/AAAAAAAAAAA/4252rscbv5M/s120/photo.jpg", "description": "Automatic update for 2026-06-14 UTC\n\nMerge-Approval-Bypass: pki-metadata-updater\nChange-Id: I384a764c775dc40ea7cbb0486e10fd010001b9b7\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7929864\nCommit-Queue: PKI Metadata Updates Bot \u003cmdb.chrome-pki-metadata-release-jobs@google.com\u003e\nBot-Commit: PKI Metadata Updates Bot \u003cmdb.chrome-pki-metadata-release-jobs@google.com\u003e\nCr-Commit-Position: refs/heads/main@{#1646503}", "email": "mdb.chrome-pki-metadata-release-jobs@google.com", "hash": "5e5aa3ba73eb33abf7460d92853bcc2f64e46dfb" }, "7935528": { "author": "manukh", "avatar": "https://lh3.googleusercontent.com/-d10XqzVQcPM/AAAAAAAAAAI/AAAAAAAAAAA/IsKJx-5pCRw/s120-p/photo.jpg", "description": "Certain user actions attempt to unelide the omnibox text. This unelision\nis correctly avoided if user input is in progress to avoid replacing the\nuser text with the original text (e.g. page URL). But it was not avoided\nwhen user match selection was in progress (e.g. user had pressed down\narrow to select the 2nd suggestion).\n\nThis didn't matter before we had on-focus suggestions, since it wasn't\npossible to have match selection in progress without the popup open; and\nit wasn't possible to have the popup open without either user input in\nprogress or on-focus suggestions.\n\nReproduces with selection changes (e.g. shift + left arrow) but not\ncursor changes (e.g. ctrl + left arrow) because\n`OmniboxViewViews::OnAfterPossibleChange()` attempts to unelide if\nselection changes.\n\nSimilarly, reproduces with shift + right arrow only if the cursor wasn't\nat the end of the text and thus resulting in a selection change.\n\nReproduces with the home key but not the end key because the home key is\nspecially handled by the omnibox view to unelide since the user may want\nto look at the start of the URL; whereas the end key does not unelide.\nThis is a different code flow than the above. Home key reproduced\nregardless of whether selection or cursor actually moved (e.g. even if\ncursor was already at start of the text). And end key did not reproduce\neven if cursor moved (e.g. even if cursor was moved left beforehand).\nThough shift + end with the left-moved cursor did reproduce via the\nabove flows.\n\nReproduced with and without webui omnibox popup. The bug lay in the part\nof the omnibox edit model that's shared by both flows.\n\nThis is the kind of edge case bug that may be re-introduced when we move\nto full webui omnibox. E.g. we might not remember to consider how\nunelideing may affect non-default suggestions.\n\nSide note, an AI coder noted early on that this bug was possibly related\nto unelision (yay). However, its attempted fix did fix the original bug,\nbut also introduced other bugs and was just logically wrong (boo). Kind\nof scary that if we just verified the behavior of the AI change without\nalready knowing how the code works, we could easily introduce bad logic\nthat would be a pain to figure out later.\n\nBug: b:508312487\nChange-Id: Iae8865fe9aff282bfe83d6f6f35567105556f537\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7935528\nReviewed-by: Justin Donnelly \u003cjdonnelly@chromium.org\u003e\nAuto-Submit: manuk hovanesian \u003cmanukh@chromium.org\u003e\nCommit-Queue: manuk hovanesian \u003cmanukh@chromium.org\u003e\nCr-Commit-Position: refs/heads/main@{#1646527}", "email": "manukh@chromium.org", "hash": "8fb4e0bbe1e63b31fb0a94ff294ebcc049333dbf" }, "7936811": { "author": "chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com", "avatar": "https://lh3.googleusercontent.com/-XWAR0jsIGjI/AAAAAAAAAAI/AAAAAAAAAAA/NEHPkbym-aA/s120-p/photo.jpg", "description": "CrOS Changes (16715.0.0-1078071 -\u003e 16716.0.0-1078088):\n- External: http://go/cros-changes/85404..85417?ext=true\n- Internal: http://go/cros-changes/119477..119494\n\nUploaded by https://ci.chromium.org/b/8679083390798009489\n\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-betty-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-brya-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-jacuzzi-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-reven-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-trogdor-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-volteer-chrome\n\nCrOS-LKGM: 16716.0.0-1078088\nMerge-Approval-Bypass: Automated LKGM update\nCr-Original-Build-Id: 8679083390798009489\nCrOS-External-Manifest-Position: 85417\nCrOS-Internal-Manifest-Position: 119494\nChange-Id: I336e33fc2e85b07fc277af452d5e49d3acd07961\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7936811\nCommit-Queue: chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com \u003cchromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com\u003e\nBot-Commit: chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com \u003cchromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com\u003e\nCr-Commit-Position: refs/heads/main@{#1646454}", "email": "chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com", "hash": "18301e1d07a7054bea3487e0eec06e1ac8bad717" }, "7937392": { "author": "chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com", "avatar": "https://lh3.googleusercontent.com/-XWAR0jsIGjI/AAAAAAAAAAI/AAAAAAAAAAA/NEHPkbym-aA/s120-p/photo.jpg", "description": "CrOS Changes (16716.0.0-1078088 -\u003e 16717.0.0-1078100):\n- External: http://go/cros-changes/85417..85426?ext=true\n- Internal: http://go/cros-changes/119494..119506\n\nUploaded by https://ci.chromium.org/b/8679038347165764721\n\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-betty-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-brya-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-jacuzzi-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-reven-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-trogdor-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-volteer-chrome\n\nCrOS-LKGM: 16717.0.0-1078100\nMerge-Approval-Bypass: Automated LKGM update\nCr-Original-Build-Id: 8679038347165764721\nCrOS-External-Manifest-Position: 85426\nCrOS-Internal-Manifest-Position: 119506\nChange-Id: Ib00ff110f9358e82475ac9f62b331274e3f77056\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7937392\nCommit-Queue: chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com \u003cchromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com\u003e\nBot-Commit: chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com \u003cchromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com\u003e\nCr-Commit-Position: refs/heads/main@{#1646498}", "email": "chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com", "hash": "695e3e871baede8dc5e5c1ed11da6a1d14739fec" }, "7937471": { "author": "Eleanor Lee", "avatar": "https://lh3.googleusercontent.com/-XxXO6wG_OmM/AAAAAAAAAAI/AAAAAAAAAAA/FrKHWUJBLxc/s120-p/photo.jpg", "description": "1. Fix the context menu anchor coordinates in AL by switching from screen coordinates to window-relative coordinates.\n\n2. Prevent context menu double-rendering on desktop workspaces by dropping long-press gestures that carry a secondary button state, addressing emulated trackpad touch frames.\n\nBug: 521982129\nChange-Id: I571bb960a1b46902bfb1505466ff1f4b0e481a78\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7937471\nReviewed-by: Gazal Agarwal \u003cagazal@google.com\u003e\nCommit-Queue: Eleanor Lee \u003celeanorlee@google.com\u003e\nCr-Commit-Position: refs/heads/main@{#1646476}", "email": "eleanorlee@google.com", "hash": "501c26d65b602fb38aa0f1589495f036e24f6c06" }, "7938248": { "author": "chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com", "avatar": "https://lh3.googleusercontent.com/-XWAR0jsIGjI/AAAAAAAAAAI/AAAAAAAAAAA/NEHPkbym-aA/s120-p/photo.jpg", "description": "CrOS Changes (16717.0.0-1078100 -\u003e 16717.0.0-1078105):\n- External: http://go/cros-changes/85426..85428?ext=true\n- Internal: http://go/cros-changes/119506..119511\n\nUploaded by https://ci.chromium.org/b/8679015504948886353\n\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-betty-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-brya-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-jacuzzi-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-reven-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-trogdor-chrome\nCQ_INCLUDE_TRYBOTS=luci.chrome.try:chromeos-volteer-chrome\n\nCrOS-LKGM: 16717.0.0-1078105\nMerge-Approval-Bypass: Automated LKGM update\nCr-Original-Build-Id: 8679015504948886353\nCrOS-External-Manifest-Position: 85428\nCrOS-Internal-Manifest-Position: 119511\nChange-Id: Idb5ca7a5adff4c7667854db1d8b1611cdc7c3dd0\nReviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/7938248\nBot-Commit: chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com \u003cchromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com\u003e\nCommit-Queue: chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com \u003cchromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com\u003e\nCr-Commit-Position: refs/heads/main@{#1646514}", "email": "chromeos-ci-prod@chromeos-bot.iam.gserviceaccount.com", "hash": "c24167a51c14e7b9bb0b92c5f8463f4ef6af1665" } }, "date": "2026-06-14", "days": 1, "generatedAt": "2026-06-15T06:19:18Z", "markdown": "## Highlights\n\n- **A11y** Chromium landed the core bidi visual caret movement algorithm and state persistence, improving cursor behavior in mixed-direction text editing. [CL 7862541](https://chromium-review.googlesource.com/c/7862541)\n- **A11y** Follow-up work now persists bidi state in `FrameSelection`, making visual caret movement stable across selection changes. [CL 7862620](https://chromium-review.googlesource.com/c/7862620)\n- **UI** The omnibox no longer clobbers text when selecting a temporary match and then moving the cursor, fixing a nasty input editing glitch. [CL 7935528](https://chromium-review.googlesource.com/c/7935528)\n- **UI** The new tab page bottom bar fixes a stale screenshot issue, cleaning up a visible regression in the IA flow. [CL 7924846](https://chromium-review.googlesource.com/c/7924846) [bug 519148302](https://issues.chromium.org/issues/519148302)\n- **Rendering** Blink introduced `HyperlinkElementUtils`, a shared utility layer for hyperlink elements that should simplify DOM link behavior going forward. [CL 7899676](https://chromium-review.googlesource.com/c/7899676)\n- **Mobile** iOS removed `AutofillAgent::_pendingAutocompleteFieldID`, cutting obsolete Autofill state and reducing complexity in field handling. [CL 7928088](https://chromium-review.googlesource.com/c/7928088) [bug 40266699](https://issues.chromium.org/issues/40266699)\n- **UI** Vertical Tabs picked up context menu spacing fixes, tightening one of the rougher edges in the newer tab strip UI. [CL 7937471](https://chromium-review.googlesource.com/c/7937471) [bug 521982129](https://issues.chromium.org/issues/521982129)\n- **Mobile** iOS added Gemini config Finch params for internal testing, laying plumbing for controlled experiments. [CL 7922184](https://chromium-review.googlesource.com/c/7922184) [bug 522712050](https://issues.chromium.org/issues/522712050)\n\n## What landed by area\n\n### Text editing and selection\n\n- **A11y** Added the core visual caret movement algorithm for bidi text so left and right navigation better matches what users see on screen. [CL 7862541](https://chromium-review.googlesource.com/c/7862541) [bug 499819853](https://issues.chromium.org/issues/499819853)\n- **A11y** Persisted bidi state in `FrameSelection`, making that new caret model survive more editing and selection transitions. [CL 7862620](https://chromium-review.googlesource.com/c/7862620) [bug 499819853](https://issues.chromium.org/issues/499819853)\n\n### Omnibox and desktop browser UI\n\n- **UI** Fixed omnibox text getting overwritten when a temporary match is selected and the cursor moves, which should make search editing less fragile. [CL 7935528](https://chromium-review.googlesource.com/c/7935528)\n- **UI** Relanded the bookmark bar auto-hide promo, bringing back a user education surface that had previously been backed out. [CL 7922377](https://chromium-review.googlesource.com/c/7922377)\n- **UI** Fixed empty space bugs in the Vertical Tabs context menu, tightening layout in that surface. [CL 7937471](https://chromium-review.googlesource.com/c/7937471) [bug 521982129](https://issues.chromium.org/issues/521982129)\n\n### New tab page and surfaces\n\n- **UI** Fixed a stale screenshot issue in the NTP bottom bar IA flow, addressing a user-visible state mismatch. [CL 7924846](https://chromium-review.googlesource.com/c/7924846) [bug 519148302](https://issues.chromium.org/issues/519148302)\n\n### Blink and DOM\n\n- **Rendering** Introduced `HyperlinkElementUtils` for hyperlink elements, a cleanup that centralizes shared link behavior in the DOM layer. [CL 7899676](https://chromium-review.googlesource.com/c/7899676)\n\n### iOS\n\n- **Mobile** Removed `AutofillAgent::_pendingAutocompleteFieldID`, trimming old Autofill bookkeeping from the iOS stack. [CL 7928088](https://chromium-review.googlesource.com/c/7928088) [bug 40266699](https://issues.chromium.org/issues/40266699)\n- **Mobile** Added Gemini config Finch flag params for internal iOS testing, setting up experiment control for upcoming work. [CL 7922184](https://chromium-review.googlesource.com/c/7922184) [bug 522712050](https://issues.chromium.org/issues/522712050)\n\n### ChromeOS and branch management\n\n- **Build** Rolled ChromeOS LKGM to `16716.0.0-1078088`, updating the known-good milestone baseline. [CL 7936811](https://chromium-review.googlesource.com/c/7936811)\n- **Build** Rolled ChromeOS LKGM to `16717.0.0-1078100`, continuing branch baseline advancement. [CL 7937392](https://chromium-review.googlesource.com/c/7937392)\n- **Build** Rolled ChromeOS LKGM to `16717.0.0-1078105`, another quick follow-up baseline update. [CL 7938248](https://chromium-review.googlesource.com/c/7938248)\n\n### Infra and sync imports\n\n- **Build** Chromium Test automation pulled in an automatic update from google3, keeping the mirrored code in sync. [CL 7929864](https://chromium-review.googlesource.com/c/7929864)\n\n## Reverts \u0026 Reapplies\n\n- **UI** Reapplied the bookmark bar auto-hide promo after an earlier rollback, suggesting the feature is back with fixes in place. [CL 7922377](https://chromium-review.googlesource.com/c/7922377)", "rollGroups": [ { "name": "Chrome Android ARM64 Orderfile", "count": 17 }, { "name": "Chrome Win64 PGO Profile", "count": 7 }, { "name": "Chrome Win32 PGO Profile", "count": 6 }, { "name": "Chrome Mac PGO Profile", "count": 5 }, { "name": "Chrome Win ARM64 PGO Profile", "count": 4 }, { "name": "Other", "count": 2 }, { "name": "compiler-rt from c556c9b6513b to bc97b63a261e", "count": 1 }, { "name": "jsoncpp from d4d072177213 to 8519b8381f3c", "count": 1 }, { "name": "clank/internal/apps from 9808c213c0b5 to be282522e173", "count": 1 }, { "name": "Perfetto Trace Processor Linux ARM", "count": 1 }, { "name": "Perfetto Trace Processor Mac ARM64", "count": 1 }, { "name": "vulkan-deps from 7bfc18a5a2d4 to 5e628ecf52c6", "count": 1 }, { "name": "WebView ARM64 Orderfile from IyTsm74FKyl69bQFZ...", "count": 1 }, { "name": "fast_float from upstream/v8.2.9 to upstream/v8.2.10", "count": 1 }, { "name": "jsoncpp from 8281fffdecb6 to 5f1f240f10a1", "count": 1 }, { "name": "WebView ARM64 Orderfile from cODJ4z5Qf05UFWiXe...", "count": 1 }, { "name": "ChromeOS Atom AFDO profile", "count": 1 }, { "name": "vulkan-deps from 5e628ecf52c6 to 5276901e0264", "count": 1 }, { "name": "Skia from 4e2c9b5e4dad to f46928e7f50c", "count": 1 }, { "name": "WebView ARM64 Orderfile from UKqXPbSTEU2WpZo3i...", "count": 1 }, { "name": "WebView ARM64 Orderfile from tIlnvSfjA9v-E6RXk...", "count": 1 }, { "name": "WebView ARM64 Orderfile from ZsJsX1mRPLFl998pD...", "count": 1 }, { "name": "Skia from c52667607242 to 4e2c9b5e4dad", "count": 1 }, { "name": "WebView ARM Orderfile from zNTrldpWgpOH7TEnH...", "count": 1 }, { "name": "Perfetto Trace Processor Linux ARM64", "count": 1 }, { "name": "Media App from ruxvnM6SHgMtIQTP5...", "count": 1 }, { "name": "Dawn from 9533d3296a2a to 5dad4a08d684", "count": 1 }, { "name": "vulkan-deps from 5276901e0264 to 5d7fd70ddbd6", "count": 1 }, { "name": "Boca App from XUye0FUo-60soMCwS...", "count": 1 }, { "name": "Chrome Android ARM32 PGO Profile", "count": 1 }, { "name": "ChromeOS Bigcore AFDO profile", "count": 1 }, { "name": "Chrome Linux PGO Profile", "count": 1 }, { "name": "Perfetto Trace Processor Win", "count": 1 }, { "name": "jsoncpp from 8519b8381f3c to 8281fffdecb6", "count": 1 }, { "name": "ios_internal from 3a997e3fe460 to db9dc5874a89", "count": 1 }, { "name": "Projector App from WaBCdSBXsz9pIwTvl...", "count": 1 }, { "name": "WebView ARM64 Orderfile from 76xAaVZoPZfXAdy11...", "count": 1 }, { "name": "WebView ARM64 Orderfile from CEAgjqQ6z8Ssn1O_L...", "count": 1 }, { "name": "Chromium Variations from sQMUMuAhgLYwWbIY6...", "count": 1 }, { "name": "libpfm4 from fbcd0c258b79 to 6870a9f00412", "count": 1 }, { "name": "ios_internal from db9dc5874a89 to a4c231c22dde", "count": 1 }, { "name": "ios_internal from a4c231c22dde to 3d896efacffb", "count": 1 }, { "name": "Help App from EPfKpwH-qMzgkF9wu...", "count": 1 }, { "name": "BoringSSL from 5124f7e67b3c to 825a42ab6426", "count": 1 }, { "name": "Boca Receiver App from 8bHpx3vpOsZ1eZ3vz...", "count": 1 } ], "since": "2026-06-14", "totals": { "commits": 93, "features": 13, "rolls": 80 }, "until": "2026-06-14" }