Uh oh!
There was an error while loading. Please reload this page.
[RN][0.74rc] Keep ES6Proxy enabled in bridgeless mode - #43539
Closed
fkgozali wants to merge 293 commits into
Closed
Conversation
Summary: Convert RNLog to Kotlin Changelog: [Internal] [Changed] - Convert RNLog to Kotlin Reviewed By: mdvacca Differential Revision: D54010172 fbshipit-source-id: eff124f094248563a2b80575d5ca9e8dae2563f7
…#43127) Summary: This adds a new warning for React Native 0.74, implementing the [RFC 0759](https://github.com/react-native-community/discussions-and-proposals/blob/nc/rnf/proposals/0759-react-native-frameworks.md#the-init-command) init command changes. - It's added inside `react-native/cli.js` to avoid warning users when actually executing `npx react-native-community/cli` commands. - The check is fairly simple: `process.argv[2] === 'init'`. The first two args are the Node bin and the actual script bin paths. - The message is sent over `console.warn` to avoid potentially mixing JSON with non-JSON output. ## Changelog: [GENERAL] [ADDED] - Warn with future command when using `npx react-native init` Pull Request resolved: #43127 Test Plan: Any command other than `init` must not warn. - `$ node ./path/to/react-native/cli.js init` - `$ node ./path/to/react-native/cli.js init --help` - Should warn with `Running: npx react-native-community/cli init`  - `$ node ./path/to/react-native/cli.js --help` - Must not warn  Reviewed By: cipolleschi Differential Revision: D54063131 Pulled By: cortinico fbshipit-source-id: c60b8b6034087b584e98b51f5bedf68a46caf44c
Summary: Fix CI build for InteropModuleRegistryTest changelog: [internal] internal Reviewed By: bvanderhoof Differential Revision: D54166759 fbshipit-source-id: 277c0a34dd4689eff8ef99afaf147efdfd36caa4
Summary: Fix Build of RNLog in CI changelog: [internal] internal Reviewed By: bvanderhoof Differential Revision: D54167011 fbshipit-source-id: 7e1ea21860b479e9fe00df5951fe266f2e342eda
Summary: Changelog: [Internal] Resyncs `react-native/debugger-frontend` from GitHub - see `rn-chrome-devtools-frontend` [changelog](react/react-native-devtools-frontend@12a45e0...9ceb0ad ). Reviewed By: huntie Differential Revision: D54107317 fbshipit-source-id: 59e670d107e4c0f30900b11df2ad423004e05997
Summary: Changelog: [Internal] Uses the capability introduced in react/react-native-devtools-frontend#4 to avoid repeating the dev server's host:port in the `ws` / `wss` parameter we pass to the Chrome DevTools frontend. This gives us more flexibility to handle port forwarding and redirects outside of `dev-middleware`. This is mostly useful in Meta's internal VS Code remoting setup, but this particular change should work equally well in open source. Reviewed By: huntie Differential Revision: D54107316 fbshipit-source-id: 68d4dbf4849ca431274bfb0dc8a4e05981bdd5b5
…ern CDP registry by default (#43101) Summary: Pull Request resolved: #43101 Extends `ReactInstanceIntegrationTest` to allow varying feature flags in tests, using gtest's parameterised tests. Exercise this in `ConsoleLogTest` to test against the modern CDP registry, for which we also needed to modify some initialisation logic to account for the fact that under the modern registry, the page is added by the host, rather than by Hermes `DecoratedRuntime`. Changelog: [Internal] Reviewed By: motiz88 Differential Revision: D53919148 fbshipit-source-id: 4eb87abf548f30b5483b819a2dadd444d1d5c80d
Summary: Changelog: [Internal] Reviewed By: GijsWeterings Differential Revision: D54166486 fbshipit-source-id: 204102267edb9d975b166ecd27b17650f554d24a
…mponent.android.js to fix CI Summary: This change renames `PopupMenuAndroidNativeComponent.js` to `PopupMenuAndroidNativeComponent.android.js`. The reason is that, without the suffix, Codegen was reading the NativeComponent spec also for iOS, generating some invalid specs and making RNTester fail. ## Changelog: [Android][Changed] - Rename `PopupMenuAndroidNativeComponent.js` to `PopupMenuAndroidNativeComponent.android.js` Reviewed By: cortinico, dmytrorykun Differential Revision: D54199736 fbshipit-source-id: 7fd67c4d38a69fe3a84c800c8ee5dcbd8c4f9a6c
Summary: This small PR fixes issue causing `AlertExample` to crash on `login-password` prompt example, as it was trying to render object in `<Text>` ## Changelog: [INTERNAL] [FIXED] - Prevent alert example from crashing Pull Request resolved: #43084 Test Plan: `login-password` prompt example in `AlertExample` doesn't crash when pressing `OK` Reviewed By: cipolleschi Differential Revision: D53964494 Pulled By: lunaleaps fbshipit-source-id: 16a0364d3d65a33956c21a68b121e6c26b41d123
Summary: Pull Request resolved: #43069 Changelog: [General][Changed] - Add info to invalid blob response error Reviewed By: christophpurrer Differential Revision: D53829400 fbshipit-source-id: 17b1dbb3cacf8368098b0b639e8d10aa9bd28d03
… a promise (#42895) Summary: `Image.getSize/getSizeWithHeaders` are still working in old fashioned "callback" way ```tsx Image.getSize(uri, function success(width,height) { }, function failure(){ } ); // undefined Image.getSizeWithHeaders(uri, headers, function success(width,height) { }, function failure(){ } ); // undefined ``` But in 2024 more developers prefer use async/await syntax for asynchronous operations So, in this PR I added support for Promise API with **backward compatibility**, modern way: ```tsx Image.getSize(uri).then(({width,height}) => { }); // Promise Image.getSizeWithHeaders(uri, headers).then(({width,height}) => { }); // Promise ``` bypass-github-export-checks ## Changelog: [GENERAL] [ADDED] - `Image.getSize/getSizeWithHeaders` method returns a promise if you don't pass a `success` callback Pull Request resolved: #42895 Test Plan: 1. ts: New test cases added in typescript tests 2. runtime: you can create a new project and put code from this PR into the next files a. `node_modules/react-native/Libraries/Image/Image.android.js` b. `node_modules/react-native/Libraries/Image/Image.ios.js` Reviewed By: javache Differential Revision: D53919431 Pulled By: cipolleschi fbshipit-source-id: 508b201e17e0ffda2e67aa5292bf9906b88d09c5
Summary: ## Changelog: [Internal] - `view.transformOrigin` prop should be initialized to "center", since this is default [per documentaion](https://reactnative.dev/docs/next/transforms#transform-origin), and it should be treated this way even if the prop is not explicitly set from JS. Reviewed By: christophpurrer Differential Revision: D54229772 fbshipit-source-id: 5212792c8dc5db6f4c17d1b2980ac2564c986cd8
Summary: This adds initial support for wide gamut (DisplayP3) colors to React Native iOS per the [RFC](react-native-community/discussions-and-proposals#738). It provides the ability to set the default color space to sRGB or DisplayP3 and provides the native code necessary to support `color()` function syntax per the [W3C CSS Color Module Level 4](https://www.w3.org/TR/css-color-4/#color-function) spec. It does _not_ yet support animations and requires additional JS code before fully supporting the `color()` function syntax. bypass-github-export-checks ## Changelog: [IOS] [ADDED] - Add basic DisplayP3 color support Pull Request resolved: #42830 Test Plan:  Follow test steps from #42831 to test support for `color()` function syntax. To globally change the default color space to DisplayP3 make the following changes to RNTester AppDelegate.mm: ```diff + #import <React/RCTConvert.h> - (BOOL)application:(UIApplication *)application didFinishLaunchingWithOptions:(NSDictionary *)launchOptions { // ... + RCTSetDefaultColorSpace(RCTColorSpaceDisplayP3); return [super application:application didFinishLaunchingWithOptions:launchOptions]; } ``` Reviewed By: javache Differential Revision: D53380407 Pulled By: cipolleschi fbshipit-source-id: 938523958f9021e8d98bdb1d4e254047e3ecdad7
Summary: This diff removes extra argument from the `extractLibrariesFromJSON` call inside `findLibrariesFromReactNativeConfig`. This should fix the iOS failurte discribed in #43204 Changelog: [iOS][Fixed] - Codegen correctly handles react-native.config.js. Reviewed By: cipolleschi Differential Revision: D54248400 fbshipit-source-id: 2ae5d0d29f49725877559a5b0edd7d59f8bdefaa
Summary: This change fixes a couple of issues within the RCTUIManager: * it calls the right method in the `super` branches (although they should neve be executed) * it invert the call order between the `_registry` and the `uiManager` to avoid extra calls into the `viewForReactTag`. ## Changelog: [Internal] - Use the right method in super and invert the order of where we search for views. ## Facebook: See S397861 and T180527210 for more information. Reviewed By: javache Differential Revision: D54246220 fbshipit-source-id: 1c7503ad3e80cf50ecc016a984ca180a19b73cc0
Summary: Changelog: [Internal] Reviewed By: cipolleschi Differential Revision: D54261247 fbshipit-source-id: 5f34372c8ccd2132f59c8369a820a5e877b2ec08
Summary: Changelog: [Internal] - Specify `access` flag in publishing nightlies Reviewed By: cipolleschi Differential Revision: D54230208 fbshipit-source-id: 7711eee62e8e01a12d6604831754d36a5cd5dc13
…n-number arguments where RootTags are expected
Summary:
Changelog: [General][Fixed] Fixed crash when passing non-numeric values where RootTag is expected to methods in native modules using codegen
## Context
Right now, if you have a native module using the codegen with a method like this:
```
someMethod(value: RootTag): void;
```
And you call it like this:
```
NativeModule.someMethod('');
```
The app crashes.
This happens because we cast the JS value to a C++ value using the method that asserts (`toNumber`) instead of the one that throws a JS error (`asNumber`).
## Changes
This fixes the crash by using `asNumber` instead of `toNumber`.
Reviewed By: RSNara
Differential Revision: D54206288
fbshipit-source-id: 9398112667e0f26edaf4f8f3b32e79fa8aafde62…hen not passing required parameters Summary: Changelog: [General][Fixed] Fixed crash when passing fewer arguments than expected in native modules using codegen ## Context Right now, if you have a native module using the codegen with a method like this: ``` someMethod(value: number): void; ``` And you call it like this: ``` NativeModule.someMethod(); ``` The app crashes. This happens because the codegen tries to cast the value to the expected type without checking if the argument is within the bounds of the arguments array. ## Changes This fixes that issue with a change in the codegen to guard against this in the generated code (see changes in the snapshot tests). Reviewed By: RSNara Differential Revision: D54206287 fbshipit-source-id: 575af462725515928f8634fccc7a9cb51ca0ce4f
Summary: ## Changelog: [Internal] - Was writing some unit tests, accidentally included `ImageProps.h` twice (once directly and once transitively) and realized that we have a handful of files without `#pragma once`. This fixes it for header files inside `ReactCommon`. Reviewed By: zeyap Differential Revision: D54258058 fbshipit-source-id: 70f4e9935304803187d1affd72ed44157b1d8fb3
Summary: Changelog: [Internal] - Changelog for 0.73.5 and 0.71.17 Pull Request resolved: #43216 Reviewed By: hezi Differential Revision: D54270560 Pulled By: lunaleaps fbshipit-source-id: 34167f5ba44c230c86896e61bff733e2be11367f
Summary: This adds support for 64 bit integer (long) values to MapBuffer. Per the wide gamut color [RFC](react-native-community/discussions-and-proposals#738) Android encodes wide gamut colors as long values so we need to update MapBuffer to support 64 bit integers as well. ## Changelog: [ANDROID] [ADDED] - Add 64 bit integer (long) value support to MapBuffer Pull Request resolved: #43030 Test Plan: I've added a test to the MapBuffer test suite. This new API is otherwise currently unused but will be used in subsequent PRs as part of wide gamut color support changes. Reviewed By: mdvacca Differential Revision: D53881809 Pulled By: NickGerleman fbshipit-source-id: 39c20b93493a2609db9f66426640ef5e97d6e1a8
…heir props Summary: This change align the `getSurfacePresenter` and `getModuleRegistry` to the iOS convention for which these should be computed properties with no `get` prefix in their name. We want to land this change and to pick it in 0.74 so we can remove the `get` versions in 0.75. ## Changelog: [iOS][Deprecated] - Deprecate `getSurfacePresenter` and `getModuleRegistry` for `surfacePresenter` and moduleRegistry` props. Reviewed By: javache Differential Revision: D54253805 fbshipit-source-id: e9ff7db744a73a3bd0f8ae1d87875e54ddd9a1a4
Summary:
This PR fixes an issue that `_logWarnIfCreateRootViewWithBridgeIsOverridden` was called in wrong place.
Assuming user overrides this method and call to `[super]`:
```objc
- (UIView *)createRootViewWithBridge:(RCTBridge *)bridge moduleName:(NSString *)moduleName initProps:(NSDictionary *)initProps {
UIView *view = [super createRootViewWithBridge:bridge moduleName:moduleName initProps:initProps];
view.backgroundColor = [UIColor redColor];
return view;
}
```
This method still wasn't called in bridgeless (and not showing the error).
Checking if user overrides this method in `appDidFinishWithLaunching` works every time
## Changelog:
[IOS] [FIXED] - Properly warn about `createRootViewWithBridge` being deprecated
Pull Request resolved: #43146
Test Plan: Check if warning is shown when message is overridden
Reviewed By: huntie
Differential Revision: D54303506
Pulled By: cipolleschi
fbshipit-source-id: cf30555c791493f28b3015a189cf93b60cace8f8Summary: All these classes are NullSafe, let's mark them as NullSafe(Local) to ensure lint detect errors in the future bypass-github-export-checks changelog: [internal] internal Reviewed By: arushikesarwani94 Differential Revision: D54027177 fbshipit-source-id: 13b4352f0b61eec69bfdcb6a3e369faa4dfcc750
Summary: All these classes are NullSafe, let's mark them as NullSafe(Local) to ensure lint detect errors in the future bypass-github-export-checks changelog: [internal] internal Reviewed By: javache Differential Revision: D54027186 fbshipit-source-id: 2cee2b598c3ef15641222018703a7d6a467ba30c
Summary: All these classes are NullSafe, let's mark them as NullSafe(Local) to ensure lint detect errors in the future bypass-github-export-checks changelog: [internal] internal Reviewed By: javache Differential Revision: D54027180 fbshipit-source-id: bf8875c06c8990172e6b449e12902691131b9cef
Summary: All these classes are NullSafe, let's mark them as NullSafe(Local) to ensure lint detect errors in the future bypass-github-export-checks changelog: [internal] internal Reviewed By: javache Differential Revision: D54027184 fbshipit-source-id: 722a7e398849f5d935894f321aa0177167eebaef
Summary: Changelog: [internal] This adds a new type of executor in AndroidExecutors to execute runnables on the UI thread. If the caller is already on the UI thread it'd call the runnable immediately. Otherwise it'd be scheduled in the UI thread to execute asynchronously. Reviewed By: huntie Differential Revision: D53941120 fbshipit-source-id: b68c7a4540be2a12df930e4e52eeb7b7a1aa91d8
Summary: Pull Request resolved: #43513 The previous `update_podfile_lock.sh` script would fail as executed from the repo root (could not locate RNTester dir). Delete this and replace with direct calls in `prepare-package-for-release.js`, which will fail script on error. {F1469216632} Changelog: [Internal] Reviewed By: cortinico Differential Revision: D54949214 fbshipit-source-id: 4f032069e803e84f835c279d01332d16787dfafc
Summary: Pull Request resolved: #43481 Update `eslint` to `8.57.0` to satisfy the peer dependency requirements of some incoming plugin updates. Changelog: [Internal] Reviewed By: huntie Differential Revision: D54898017 fbshipit-source-id: 39a8ac0f9653a104277a7f0054bb7e1fe3ec54f4
Summary: Pull Request resolved: #43516 As titled. This seems dangerous — removing with the motivation that we'd prefer this script to fail during execution than to succeed in publishing `9999`. Changelog: [Internal] Reviewed By: cortinico Differential Revision: D54956661 fbshipit-source-id: 23f8d49abd300385dde74871b6d2492ef63f058e
Summary: Changelog: [Internal] bypass-github-export-checks Reviewed By: huntie Differential Revision: D54957923 fbshipit-source-id: 2178f2715e9c924ee58ac39b728279c914d4891d
Summary: Pull Request resolved: #43514 Changelog: [Internal] The navbar active/inactive colors were fundamentally broken on dark theme. Reviewed By: rubennorte Differential Revision: D54952629 fbshipit-source-id: 9cfd7d4eb904c7b6a8d7e1e900b4ff076b8aba12
Summary: This is causing Twilight Android app to hang and crash on start Original commit changeset: 3f24271405f6 Original Phabricator Diff: D54905564 Reviewed By: sammy-SC Differential Revision: D54960053 fbshipit-source-id: 5c1063f11ae1314e71288905eb6d8da32ddbfadd
…#43510) Summary: Pull Request resolved: #43510 changelog: [internal] - use designated initializers. - make TextInputMetrics and KeyPressMetrics structs. - move TextInputMetrics and KeyPressMetrics inside of TextInputEventEmitter. - Utilise RCTSizeFromCGSize, RCTEdgeInsetsFromUIEdgeInsets and RCTPointFromCGPoint when creating TextInputMetrics. Reviewed By: rshest Differential Revision: D54895928 fbshipit-source-id: 4afd49ffa77c5afafc794808aac7f11bb8213944
Summary: Pull Request resolved: #43518 This is a minimum approach to achieve a **single-command publish flow** for React Native, unifying the previous `yarn bump-all-updated-packages` and `yarn trigger-react-native-release` workflow entry points. This diff aims to change as little as possible to achieve the above — introducing a new job that merges operations to create the versioning commit. The triggered publish jobs are unchanged. In future, we may follow this change with further simplifications down the workflow tree. **Key changes** - Adds a new CircleCI workflow, `prepare_release_new`, which versions **all packages** and writes a single release commit. - This replaces `yarn bump-all-updated-packages`, now implemented with the newer `set-version` script. - Wires this up as an experiment within `trigger-react-native-release.js`, conditionally running the new workflow when `--use-new-workflow` is passed. **Not changed** - The single release commit written will continue to trigger both of the existing CI workflows on push (`publish_release` and `publish_bumped_packages`), which are unchanged. - The commit summary now includes the `#publish-packages-to-npm` marker, in order to trigger `publish_bumped_packages`. - Usage: Release Crew members will continue to use the existing local script entry point (as [documented in the releases repo](https://github.com/reactwg/react-native-releases/blob/main/docs/guide-release-process.md#step-7-publish-react-native)), with the opt in flag. ``` yarn trigger-react-native-release --use-new-workflow [...args] ``` After we're happy with the E2E behaviour of this workflow in the next 0.74 RC, I will follow up by dropping the `--use-new-workflow` flag and removing the old scripts (T182533699). Changelog: [Internal] Reviewed By: cortinico Differential Revision: D54956345 fbshipit-source-id: 35fd7af8f3e60a39507b5d978ccd97472bf03ddb
Summary: Pull Request resolved: #43507 Changelog: [General][Changed] - sync React renderers to 18.3.0-canary-9372c6311-20240315 Syncs React renderers to react/react@9372c63 which is canary for 18.3.0-canary-9372c6311-20240315. This includes the necessary changes to enable the use of microtasks for scheduling in Fabric. Reviewed By: yungsters Differential Revision: D54947212 fbshipit-source-id: 8fd5def5107d77e6a248f653a9d0260b392fab6b
Summary: Pull Request resolved: #43396 Changelog: [General][Changed] Enabled new event loop behavior when bridgeless (new architecture) is enabled. Reviewed By: rshest Differential Revision: D54682678 fbshipit-source-id: ff8c45bc1caae0e9182aa94d915d7b6f9427caf9
Summary: Pull Request resolved: #43503 Changelog: [internal] ## Context In the experiments to enable mount hooks on Android we saw some crashes that we couldn't reproduce or pinpoint accurately. We ran another experiment excluding part of the code in one of the variants, and we found that the problem was in this block (only crashes when `skipMountHookNotifications` is false): https://github.com/facebook/react-native/blob/121b26184acbb77ff4f2360647cb322ff560b145/packages/react-native/ReactCommon/react/renderer/uimanager/UIManager.cpp#L726-L734 Looking more closely at the mounting coordinator code, I realized that some of the methods are not thread-safe, which is likely causing these issues. ~~We're probably only seeing these issues on Android because we have a push model there (we call `pullTransaction` from whatever thread we're committing to, JS thread or Fabric background thread) whereas in the rest of platforms we have a pull model and we always access call `pullTransaction` from the main thread, as we do to report mounts.~~ This is probably fine because both cases are protected by a mutex when accessing through `ShadowTreeRegistry::visit`. But there's a case that doesn't go through it that could be causing the issues: prerendering: https://github.com/facebook/react-native/blob/121b26184acbb77ff4f2360647cb322ff560b145/packages/react-native/ReactCommon/react/renderer/scheduler/SurfaceHandler.cpp#L289 ## Changes 1) Make `getBaseRevision` return a copy of the revision rather than a reference. 2) Make all methods that access `baseRevision_` and `lastRevision_` thread-safe. Reviewed By: javache Differential Revision: D54945100 fbshipit-source-id: d8b211137d0eac02a5814cc6c376c22733290eab
Summary: Pull Request resolved: #43504 Changelog: [internal] We identified the issue in the experiments already, so we can remove this flag. Reviewed By: sammy-SC Differential Revision: D54945099 fbshipit-source-id: 4d547569eb3bbfd011f5d6894d87bfa542cac07b
Summary: ### Context - Since RN 0.73, in the jest.setup file, methods of the Image module (Image.getSize, Image.resolveAssetSource...) are mocked on the **JS side** (introduced in #36996) - It causes issues like #41907 : `Image.resolveAssetSource` returns nothing in test env with the new JS mock, when some test relies on it. - On my project, it broke the snapshots : the URL of images disappeared. I use `react-native-fast-image` which uses `Image.resolveAssetSource` to compute URLs. - I first opened a PR to fix exclusively Image.resolveAssetSource: #41957. I will close it to focus on this new one. - As suggested by ryancat and idrissakhi, it should be better to return to the previous mock, where no method is mocked on the JS side, and we can trust the actual JS to work in test. This is what this PR intends to do. ### Content Along fixing the Image module mock in jest.setup, this PR : - adds unit test on each one of the methods, ensuring they have a consistent behavior even when the module is mocked. - adds 3 missing native mocks for `NativeImageLoader`: `prefetchImageWithMetadata`, `getSizeWithHeaders` & `queryCache`. After this PR, no method from NativeImageLoader remains unmocked. ## Changelog: [GENERAL][FIXED] - fix jest setup for Image methods (resolveAssetSource, getSize, prefetch, queryCache) Pull Request resolved: #43497 Test Plan: See exhaustive unit tests in PR. You can re-use the mock with all the methods mocked and see how the new unit tests fail. I also patched those changes on my project: my snapshot did have their URL back (see demonstrative screenshots in my original PR: #41957 - NB; fixed mock was different but result was the same -> those screenshots cover only two cases, but anyway they illustrate well the case!) Reviewed By: ryancat Differential Revision: D54959063 Pulled By: tdn120 fbshipit-source-id: 837266bd6991eb8292d9f6af1774e897ac7a8890
) Summary: Pull Request resolved: #43385 `rncore` and `FBReactNativeComponentSpec` contain the same symbols, which leads to conflicts when we try to merge them into a single shared library. Cleanup the duplication and standardize on `FBReactNativeComponentSpec` everywhere. I've left the Android OSS targets as is, to avoid breaking deps. Changelog: [Internal] Reviewed By: cortinico, dmytrorykun Differential Revision: D54630694 fbshipit-source-id: 75cb961ded9fd75508755c0530e29409fef801cf
… to v27 (#43406) Summary: This PR updates `typescript-eslint/eslint-plugin` and `typescript-eslint/parser` to `v7` and `eslint-plugin-jest` to `v27`, removing any dependencies on `typescript-eslint` `v6`, allowing projects using `react-native/eslint-config` to safely update to `typescript-eslint` `v7` without having to worry about duplicate major versions installed ## Changelog: - [General] [Changed]: Updated `eslint-plugin-jest` to `v27` - [General] [Changed]: Updated `typescript-eslint` monorepo to `v7` Pull Request resolved: #43406 Test Plan: `yarn run lint` executed locally successfully Reviewed By: robhogan Differential Revision: D54749676 Pulled By: tdn120 fbshipit-source-id: f6fae92fc95333e28b36a3d2bd8470c8869d38bc
Summary:
Minor fix to package.json which newer version of npm warn about when publishing, after running `npm pkg fix -ws` on the workspace.
{F1470070110}
## Changelog: [Internal] npm pkg fix -ws
Pull Request resolved: #43519
Test Plan: eyescloseddog
Reviewed By: cortinico
Differential Revision: D55012872
Pulled By: blakef
fbshipit-source-id: ff3c63a3eefaf56d369219a3d4b32d44d6d842c9…ComponentSpec Differential Revision: D54630694 Original commit changeset: 75cb961ded9f Original Phabricator Diff: D54630694 fbshipit-source-id: 50832c934e5bcc1a92456d949995c05a33767a92
…43524) Summary: The [Windows fix](cezaraugusto/chromium-edge-launcher#1) was merged and published. We no longer need to use the fork. ## Changelog: [INTERNAL] [FIXED] - Fix experimental debugger launch flow with Edge on Windows Pull Request resolved: #43524 Test Plan: n/a Reviewed By: robhogan Differential Revision: D55013623 Pulled By: motiz88 fbshipit-source-id: bff2aa2801dd0dcdd6975dca0a2ec2aa9864ff6f
Summary: Pull Request resolved: #43520 Refactor ReactDelegate to have a private `getDevSupportManager()` that can also be re-used by `reload()` This method conditionally provides the correct DevSupportManager in cases of Bridge & Bridgeless Changelog: [Internal] internal Reviewed By: cortinico Differential Revision: D54967130 fbshipit-source-id: 37d585de33a50b98d01803d3080c5693a8c494b9
Summary: Pull Request resolved: #43530 Fixes and changes following D54956345, encountered during the release process for 0.74 RC4 today. Changelog: [Internal] Reviewed By: cortinico Differential Revision: D55017872 fbshipit-source-id: 616b387088db00c6f076f4571b4ab1541467361c
…s fix was merged Differential Revision: D55013623 Original commit changeset: bff2aa2801dd Original Phabricator Diff: D55013623 fbshipit-source-id: d7f19068327b5cf427750fbcae01d3c10f9c89b5
Summary: The Hermes RuntimeConfig for bridgeless accidentally force-disabled ES6Proxy, resulting in #43523 To fix this in the 0.74rc, let's remove the incorrect override. Note that this is only for the RC branch, the fix in main will be done separately (to use proper ReactNativeFeatureFlags). To test using RNTester, add the following change: ``` diff --git a/packages/rn-tester/js/RNTesterAppShared.js b/packages/rn-tester/js/RNTesterAppShared.js index 87cb6b6..f2512d09c5a 100644 --- a/packages/rn-tester/js/RNTesterAppShared.js +++ b/packages/rn-tester/js/RNTesterAppShared.js @@ -50,6 +50,8 @@ const RNTesterApp = ({ ); const colorScheme = useColorScheme(); + new Proxy({}, {}); + const { activeModuleKey, activeModuleTitle, ``` Before this change, RNTester will get an error at start-up. After, the app loads correctly. Changelog: [General][Fixed] Correctly keep ES6Proxy for bridgeless mode Differential Revision: D55045780
fkgozali
commented
Mar 19, 2024
ContributorAuthor
Sorry, wrong PR, will fix |
analysis-bot
commented
Mar 19, 2024
Base commit: ff5e1a6 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
19 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary:
Note: this is a copy of #43538 but targeted towards 0.74-stable branch directly.
The Hermes RuntimeConfig for bridgeless accidentally force-disabled ES6Proxy, resulting in #43523
To fix this in the 0.74rc, let's remove the incorrect override. Note that this is only for the RC branch, the fix in main will be done separately (to use proper ReactNativeFeatureFlags).
Changelog:
[General][Fixed] Correctly keep ES6Proxy for bridgeless mode
Test Plan:
To test using RNTester, add the following change:
Before this change, RNTester will get an error at start-up. After, the app loads correctly.