[pull] canary from vercel:canary #188

### What? Copy the benchmark to the new backend Closes PACK-3417

### Why? PR #72311 bumped the repo's TS to 5.6.3 but missed out on the packages that added TypeScript to their own at #71256.

### Why? `next/font` depends on the Google Fonts metrics from the `@capsizecss/metrics`. Added support for Geist at seek-oss/capsize#217, so we update version. ### How? - Bump `@capsizecss/metrics` to 3.4.0 - Update the test to match the metrics. #### Bonus - Modified the test to use `.toMatchInlineSnapshot` to identify and update future changes more easily. ### RFC The `@capsizecss` team is planning to [automate the updating process](seek-oss/capsize#217 (review)). Should we continue to depend on them or replace them with our custom script to fetch Google Font API (requires API key)? x-ref: #47115 x-ref: [Twitter Feedback](https://x.com/illyism/status/1855500917589602706)

The option was always set to `true` so we can safely remove it. The experimental `serverActions` flag was removed a long time ago in #57145.

Observable in `pnpm next dev test/development/app-dir/dynamic-io-dev-errors/` and `/no-accessed-data`. It only affects the bottom stacks in that particular stack so assertions are not viable yet until the full stack is properly sourcemapped.

### Why? Following up on #72752, the ESLint fix suggestion for `react/no-unescaped-entities` rule was added at jsx-eslint/eslint-plugin-react#3831, released on [v7.37.0](https://github.com/jsx-eslint/eslint-plugin-react/releases/tag/v7.37.0). Therefore, we bump the `eslint-plugin-react` version to `^7.37.0`.

Tests have been inconsistent due to flaky build-time failures caused by AMP validation errors: https://github.com/vercel/next.js/actions/runs/11807856104/attempts/1. We’re updating the library version with the expectation that this will resolve the issue. ``` info: undefined Generating static pages (9/37) Generating static pages (18/37) Generating static pages (27/37) Error occurred prerendering page "/amp-hybrid". Read more: https://nextjs.org/docs/messages/prerender-error AssertionError: Assertion failed: WebAssembly is uninitialized at new module$contents$goog$asserts_AssertionError (evalmachine.<anonymous>:102:1441) at module$contents$goog$asserts_doAssertFailure (evalmachine.<anonymous>:103:354) at goog.asserts.assertExists (evalmachine.<anonymous>:104:374) at Object.module$contents$amp$validator_validateString [as validateString] (evalmachine.<anonymous>:2238:108) at Validator.validateString (/tmp/next-install-89c7ccfd9a04886e2e88ccffb9ad709b236d4b8c8398fcdbb34c9c897118b9e9/node_modules/.pnpm/next@file+..+next-repo-99a8e0af0510a7a2958e71c9affe3d5d83d452a304c40ea8b352506d74651d57+packa_6sb2ai4rhstsynpbl53dxgpt7e/node_modules/next/dist/compiled/amphtml-validator/index.js:17:2057) at validateAmp (/tmp/next-install-89c7ccfd9a04886e2e88ccffb9ad709b236d4b8c8398fcdbb34c9c897118b9e9/node_modules/.pnpm/next@file+..+next-repo-99a8e0af0510a7a2958e71c9affe3d5d83d452a304c40ea8b352506d74651d57+packa_6sb2ai4rhstsynpbl53dxgpt7e/node_modules/next/dist/export/routes/pages.js:100:34) at process.processTicksAndRejections (node:internal/process/task_queues:95:5) at async exportPagesPage (/tmp/next-install-89c7ccfd9a04886e2e88ccffb9ad709b236d4b8c8398fcdbb34c9c897118b9e9/node_modules/.pnpm/next@file+..+next-repo-99a8e0af0510a7a2958e71c9affe3d5d83d452a304c40ea8b352506d74651d57+packa_6sb2ai4rhstsynpbl53dxgpt7e/node_modules/next/dist/export/routes/pages.js:134:17) at async Span.traceAsyncFn (/tmp/next-install-89c7ccfd9a04886e2e88ccffb9ad709b236d4b8c8398fcdbb34c9c897118b9e9/node_modules/.pnpm/next@file+..+next-repo-99a8e0af0510a7a2958e71c9affe3d5d83d452a304c40ea8b352506d74651d57+packa_6sb2ai4rhstsynpbl53dxgpt7e/node_modules/next/dist/trace/trace.js:153:20) at async exportPage (/tmp/next-install-89c7ccfd9a04886e2e88ccffb9ad709b236d4b8c8398fcdbb34c9c897118b9e9/node_modules/.pnpm/next@file+..+next-repo-99a8e0af0510a7a2958e71c9affe3d5d83d452a304c40ea8b352506d74651d57+packa_6sb2ai4rhstsynpbl53dxgpt7e/node_modules/next/dist/export/worker.js:336:18) ``` Note that this failure breaks the production integration test almost 10% of times. ![CleanShot 2024-11-12 at 23 15 11@2x](https://github.com/user-attachments/assets/89ad041c-22b1-47ad-aa56-58ea3a269811)

### Why? ESLint v8 is deprecated. Also, it's deprecated dependencies spam npm warning. ![image](https://github.com/user-attachments/assets/856a70a5-3185-44d6-9fa9-c773f898dc25) Closes NDX-208

#72641) This PR adds a new example that shows a more complex use case where you don't want browser to know about the upstream image server so instead you can proxy through a serverless function. Some use cases might be checking for an IP address, user agent, etc.

### Why? As Google Font added Geist, replace the CNA local font with`next/font/google`. Added `subsets` value as the font will be preloaded. Removed`weight` value as: > An array of weight values if the font is not a [variable google font](https://fonts.google.com/variablefonts). It applies to next/font/google only. x-ref: https://nextjs.org/docs/app/api-reference/components/font#subsets x-ref: https://nextjs.org/docs/app/api-reference/components/font#weight Closes NDX-462

### What? Use proper logic to depend on the side-effect-import node from the import-binding nodes in **graph construction phase**. ### Why? We previously used a workaround based on a hash map with the module specifier as the key, but as I'm going to optimize the graph, we need to store every dependency information in the graph so that the optimizer can use information. ### How? Closes PACK-3444

`redirect` in server actions has handling to support relative redirects (eg `redirect('./foo')`) but it doesn't handle a multi-level relative path, eg `redirect('../foo')`. This used to work by accident because the invalid URL would be handled as an error and then would trigger an MPA navigation. In Next 15, it surfaces an actual error because we are less tolerant to them. Fixes #72765 Closes NDX-479

## What does this PR do? According to [Installation docs](https://nextjs.org/docs/getting-started/installation#automatic-installation), [create-next-app](https://nextjs.org/docs/app/api-reference/cli/create-next-app) messages are below ![image](https://github.com/user-attachments/assets/bb20d031-a8f0-4306-8c41-d3341fa8c899) However, the real messages are ![image](https://github.com/user-attachments/assets/828f7e15-8284-4b97-a62f-d99867aede09) Unlike the `src/`, `next dev` and `@/*` don't have single quotation. So I add apostrophes to create-next-app message. ## How did you verify your code works? I have tested local build package. ![image](https://github.com/user-attachments/assets/f358cf89-8bf1-48aa-97f9-2f097e4ef753) Co-authored-by: JJ Kasper <[email protected]>

We run flake detection on PRs to help catch flaky tests before they hit canary, but we currently only run flaky test detection in Webpack and not Turbopack. This is problematic because there's sufficient forking logic in tests for the different bundlers & we've seen a high likelihood of a test to flake more frequently in one bundler vs another due to things like differing compilation speeds and differing implementation of core features. This will add some additional time to this CI job to account for needing to run the tests 6 times, but will hopefully catch things before they become a problem later. I think we could also consider reducing the number of times per bundler to be 2 if it turns out to be adding too much time. This updates the flake detection job to run the tests multiple times in Turbopack as well. CI run: https://github.com/vercel/next.js/actions/runs/11822662639/job/32940264160?pr=72773

## Summary Follow up #72636. Bump 13 → 15 version in the description. ### Adding or Updating Examples - [x] The "examples guidelines" are followed from our contributing doc https://github.com/vercel/next.js/blob/canary/contributing/examples/adding-examples.md - [x] Make sure the linting passes by running `pnpm build && pnpm lint`. See https://github.com/vercel/next.js/blob/canary/contributing/repository/linting.md --------- Co-authored-by: samcx <[email protected]>

)

## Description I have corrected some mistakes according to the [contribution guide](https://github.com/vercel/next.js/blob/canary/contributing/examples/adding-examples.md). 1. Update `create-next-app` script 2. Remove redundant new line, space and double colon 3. `readme.md` → `README.md` ### Adding or Updating Examples - [x] The "examples guidelines" are followed from our contributing doc https://github.com/vercel/next.js/blob/canary/contributing/examples/adding-examples.md - [x] Make sure the linting passes by running `pnpm build && pnpm lint`. See https://github.com/vercel/next.js/blob/canary/contributing/repository/linting.md --------- Co-authored-by: Sam Ko <[email protected]>

### What? Follow-up from review comments

## Summary Sync [`04-functions`](https://github.com/vercel/next.js/tree/canary/docs/01-app/03-api-reference/04-functions) file convention. ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide  Co-authored-by: Sam Ko <[email protected]>

### What? Tailwind CSS [supports an ESM config](https://tailwindcss.com/docs/configuration#using-esm-or-type-script) file. This PR changes `tailwind.config.js` to `tailwind.config.mjs`. ### Why? 1. To convert one more file to ESM 2. To use the modern format 3. To follow other similar migrations that have taken place in the Next.js codebase (e.g. `next.config.mjs`) ### How? 1. Change file extensions from `.js` to `.mjs` 2. Change module format from CommonJS to ESM Co-authored-by: Sam Ko <[email protected]>

### What? fix some bugs when counting collectibles

When `next dev` is started with `-p 0` or `PORT=0` – which is probably unusual in userland, but is the default for our tests – a random port is selected for the dev server. If the user, after starting the dev server, then changes the `next.config.js` file, the dev server detects that and automatically restarts. For the restart, the original port config is used, resulting in a random port being selected again. This will most likely be a different port than was selected before, and so reloads in the browser won't work. It also results in `ERR_CONNECTION_REFUSED` errors in our tests that use `patchFile` to change the `next.config.js` content. To fix that issue, we're now storing the automatically selected port when the dev server has started initially, and are re-using it for any subsequent automatic restarts.

…ed (#72781) Using a build error instead of a runtime error allows us to fail early, and show a proper error source in the terminal and in the dev overlay. <img width="777" alt="Screenshot 2024-11-13 at 22 19 45" src="https://github.com/user-attachments/assets/d0ee3c69-71f5-4aa6-8c0a-879217f66930">

## Why? There is a type issue with one of the form actions. ``` app/(auth-pages)/sign-up/page.tsx:42:25 - error TS2322: Type '(formData: FormData) => Promise<{ error: string; }>' is not assignable to type 'string | ((formData: FormData) => void | Promise<void>) | undefined'. Type '(formData: FormData) => Promise<{ error: string; }>' is not assignable to type '(formData: FormData) => void | Promise<void>'. Type 'Promise<{ error: string; }>' is not assignable to type 'void | Promise<void>'. Type 'Promise<{ error: string; }>' is not assignable to type 'Promise<void>'. Type '{ error: string; }' is not assignable to type 'void'. 42 <SubmitButton formAction={signUpAction} pendingText="Signing up..."> ``` - x-ref: #72778

### What? When a root task is disposed we need to flag it as inactive. This is e.g. the case when closing a WebSocket, which closes the subscriptions. Not disposing causes unnecessary work which we want to avoid.

### Why? This PR silences `legacy-js-api` warning enabled from PR #70067 by updating the `sass-loader` to v15. Will follow up with adding an option to enable the sass-loader v16 by PR #72423. x-ref: #71638

Migrated every instance of `sandbox` to use the `using` keyword so that the `cleanup` will be automatically performed when the execution leaves the scope.

) ### What? Merge a dependency into the chunk if the current node is the only dependant of the dependency. i.e. Optimize for cases where only one **dependant** exists for a specific node. ### Why? If a node has a single incoming node (let's call it `source`), it means that it will always be included if the source node is included and not if the source node is not included. So the source and node act as a group, and this PR makes them a real group to reduce number of module fragments. Before: <img width="1039" alt="image" src="https://github.com/user-attachments/assets/24321064-fe34-45c6-bf02-81e482816741"> After: <img width="1039" alt="image" src="https://github.com/user-attachments/assets/4709a1f9-097c-4771-b554-043c01c980db"> ### benches/heavy-npm-deps - 14862 modules => 9263 modules - With tree shaking on canary: `144.13s user 14.13s system 910% cpu 17.373 total` - With tree shaking and this PR: `124.60s user 12.41s system 878% cpu 15.601 total` - Without tree shaking: `47.41s user 4.75s system 553% cpu 9.416 total` ### How? Closes PACK-3441

Since prefetches include only static data, and are generated during the server prerender, we can actually lift all the tokens to the route tree prefetch. It's only once we start supporting dynamic conditional routes that the tokens will need to provided by the layout. This is nice because it means we can parallelize the segment prefetches after all.

#72763)

@unstubbable

Discussion with @unstubbable was that he didn't know if he should use `String` or `RcStr`. This should clarify what people should use, why, and how to construct the `RcStr`. ![Screenshot 2024-11-14 at 15-14-48 RcStr in turbo_tasks - Rust.png](https://graphite-user-uploaded-assets-prod.s3.amazonaws.com/HAZVitxRNnZz8QMiPn4a/ad9aeb14-16f5-49d4-8433-fb25d4d1214c.png)

@samcx

…ks-fetch (#72526) Noticed this mess while working with @samcx on #72442 where we had to run: ``` cargo nextest r --features next-core/native-tls -p next-core ``` This PR makes `turbo-tasks-fetch` responsible for setting the `reqwest` TLS features, rather than trying to pass that feature flag down. After this PR, running tests on `next-core` works with just: ``` cargo nextest r -p next-core ``` ## Why? - The feature flag logic was getting duplicated in multiple top-level targets. - Packages that depended on `turbo-tasks-fetch` (directly *or transitively*) had to specify a default feature and make sure that default feature was disabled in the workspace's dependency list so that `cargo test -p package-name` would work - ... or you had to specify a flag when compiling, like `--features next-core/native-tls`. Technically this change gives us less flexibility. We can't produce multiple binaries for the same platform target with different TLS stacks. But I don't think we would've ever used that flexibility anyways.

`dynamic_embed_contents` is off by default, so i guess we missed some type errors when refactoring stuff to `RcStr`

### What? Currently when there are multiple instances of the same error on different lines of the same source file it shows a random one of these errors. This fixes that by sorting the issues also based on the source.

…2764)

### What? Follow-up for #72564 ### Why? We need to refactor this for future works. ### How? --------- Co-authored-by: Benjamin Woodruff <[email protected]>

… types (#72823) `Vc<T>` had a type bound that `T: Send + ?Sized`. The general recommendation is to remove all type bounds from the struct that aren't necessary: - https://rust-lang.github.io/api-guidelines/future-proofing.html#c-struct-bounds - rust-lang/api-guidelines#6 - rust-lang/rust-clippy#1689 The reasoning is that type bounds on structs are "infectious". Any generic function, trait, or struct referring to `Vc<T>` was required to add `T: Send` bounds. *Sidenote: The [`implied_bounds` feature](https://rust-lang.github.io/rfcs/2089-implied-bounds.html) might mitigate some of this if ever stabilized.* Removing the `T: Send` type bound from `Vc<T>` means that there's less places where we need to specify it. This pattern can be seen in many parts of the stdlib. For example, `Arc<T>` doesn't require that `T: Send + Sync`, though in reality to do anything useful with it, you need `T: Send + Sync`. ## Is this safe? Yes, bounds are checked during cell construction, and the lower-level APIs used for creating cells require `Send + Sync`, so this would be hard to mess up without explicitly unsafe code. Also, `Vc<T>` also technically requires that `T: Sync`, but that wasn't enforced in the struct definition (only during cell construction). We did just fine without that bound on the struct. ## Why are you leaving `?Sized`? There's an implicit `Sized` bound on type parameters in struct definitions unless explicitly specified otherwise with `?Sized`. Right now this bound isn't used. We use a box, e.g. `Vc<Box<dyn Foo>>`, but in the future we might be able to drop that intermediate box from the type signature. I'm leaving the `?Sized` bound in place (and adding it in a few places where it was missing) in hopes of that.

…2607) ### What? Disable CSS Modules renaming for containers. ### Why? We should match the behavior of webpack mode. See: parcel-bundler/lightningcss#835 ### How? - Closes PACK-3315 - Closes #71233

### What? Cleanup logic. - Remove duplicate imports. - Rename `exports` to `outputs` as it's more intuitive. - Do not track the `ix` of each exports, as it's not used. ### Why? ### How?

### What? We need to notify the in progress cells when task has finished or cells are updated Also fixes a off-by-one bug with max cell indicies when reading

### What? The `data` variant for pages and the `rsc` variant for app router (for change notification purposes) causes some conflicting writes in the next-rs test case. This causes some instability since we are invalidating in a loop.

#72811) This improves maintainability and readability by decluttering the server actions transforms implementation, dedupes a few error messages, and consistently uses periods and trailing newlines for all error messages. The same approach is already used for the React Server Components transforms.

### Why? We are renaming `revalidate(Tag|Path)` to `expire(Tag|Path)`. Closes NDX-482 --------- Co-authored-by: Zack Tanner <[email protected]>

Hehe. `flightData` is either an array or a string. There's a loop that's meant to iterate over `flightData` in the case where it's an array, but we neglected to check whether it was an array or a string, and since strings are iterable types, the type system did not complain. The check that should have been before the loop was instead inside the loop... and it happened to work anyway, because when you iterate over a string, each item is a character, and characters have the type "string". So the code works despite being nonsensical. In conclusion, lmfao.

This error doesn't point to any particularly meaningful line of code to the user, we just use it to have a well-known error message for when a feature should only be used in canary. This hides the stack trace so the error is surfaced more clearly.

Updates the client PPR implementation to gracefully handle missing CacheNodeSeedData when constructing a prerender tree. Currently this never happens because we always block until we receive a server response. However, in the new Segment Cache, we will not bother to wait if a URL has not already been prefetched, since it's usually faster to go straight to the dynamic request.

Wraps the `prefetch` API in a feature check for the Segment Cache, and forwards the call to a stub for the new implementation. Unlike the old implementation, the Segment Cache doesn't store its data in the router reducer state; it writes into a global mutable cache. So we don't need to dispatch a router action. Since the Segment Cache isn't actually implemented yet, this effectively disables prefetching when the experimental flag is enabled. There's some validation that we do for prefetch URLs that I extracted into a shared function. (For example, we only prefetch same-origin URLs, and we don't prefetch anything in development.)

…72865) In the Segment Cache implementation, I won't be using the main fetchFromServerResponse function because the protocol for per-segment prefetches is different. But there's a lot of logic (e.g. setting headers, decoding the Flight stream, validating the response) that can be shared. So I've extracted the shared bits into separate functions.

…72858) The `RcStr` type doesn't use anything in `turbo-tasks`, so there's no reason it can't be a separate crate that doesn't depend on `turbo-tasks`. This will let it be used in wasm swc plugins, where turbo-tasks does not currently compile. Imports were automatically rewritten with: ``` # Fix imports: sg run --lang rust --pattern 'use $CRATE::{$$$PRE, RcStr};' --rewrite 'use $CRATE::{$$$PRE}; use turbo_rcstr::RcStr;' --update-all sg run --lang rust --pattern 'use $CRATE::{RcStr, $$$POST};' --rewrite 'use $CRATE::{$$$POST}; use turbo_rcstr::RcStr;' --update-all sg run --lang rust --pattern 'use $CRATE::{$$$PRE, RcStr, $$$POST};' --rewrite 'use $CRATE::{$$$PRE, $$$POST}; use turbo_rcstr::RcStr;' --update-all sg run --lang rust --pattern 'use $CRATE::RcStr;' --rewrite 'use turbo_rcstr::RcStr;' --update-all cargo fmt # Fix strings in comments: fastmod crate::RcStr turbo_rcstr::RcStr fastmod turbo_tasks::RcStr turbo_rcstr::RcStr ``` I had to manually fix the `Cargo.toml` files based on the compilation errors.

In the current navigation implementation, a partially dynamic navigation always does two separate requests: one for the static data, and one for the dynamic data. Typically the static data is prefetched before the navigation begins, but even in the case where it is not, the current implementation will still fetch it first. It then wait to send a dynamic request until the first chunk is received from the prefetch response, leading to an unfortunate request waterfall. In the Segment Cache implementation, our plan is to never block a navigation on prefetch data that isn't already populated in the cache. Instead, in the case of a cache miss, we'll immediately start a dynamic navigation and rely on the fact that the first thing the dynamic response sends is the static PPR shell of the target page. Because we'll always have this as a fallback behavior for cache misses, it's a good starting point for the Segment Cache implementation. Then we can start incrementally adding more and more features until we've eventually reached/surpassed parity with the current implementation. --- To avoid duplication of logic, I've chosen to model cache misses as a special case of the normal static + dynamic flow. We can pretend that the route tree returned by the dynamic request is, in fact, the result of a prefetch. Then we use that same server response to write data into the CacheNode tree. So it's the same flow as the "happy path", except we use a single server response for both stages.

This breaks out the type overriding note from a small good to know section to it's own section as it's fairly important and easy to miss currently. Also the minimal note isn't getting picked up in our search currently. x-ref: [slack thread](https://vercel.slack.com/archives/C0676QZBWKS/p1731602395153929)

) ### What? Use `ResolvedVc` for struct fields in `turbopack-ecmascript` crate ### Why? Follow-up of #72200, #72382, #72320 ### How? Closes PACK-3438 --------- Co-authored-by: Benjamin Woodruff <[email protected]>

### What? Previously, we had a workaround while consuming the graph and converting them to `Module`s. ### Why? Everything should be stored in the dependency graph because we are going to optimize it using graph node traversal. ### How?

@type

When a `"use cache"` directive with a custom cache kind is used, e.g. `"use cache: custom"`, a cache handler with the same name must be specified in the Next.js config: ```js /** * @type {import('next').NextConfig} */ const nextConfig = { experimental: { dynamicIO: true, cacheHandlers: { custom: require.resolve('path/to/custom/cache/handler'), }, }, } module.exports = nextConfig ``` If this is not the case, we emit a build error with an error message that explains this requirement. <img width="795" alt="Screenshot 2024-11-14 at 23 30 01" src="https://github.com/user-attachments/assets/bd138c97-608f-42ce-abb2-edb3e44edc3f"> When we'll get a docs page for this experimental config, we will add the usual "Read more: ..." hint as well. --------- Co-authored-by: Benjamin Woodruff <[email protected]> Co-authored-by: Janka Uryga <[email protected]>

) ### Why? Before we just wrote the files in `FileSystem::write()`, but that caused some issues when two `FileSystem::write()` tasks are active (or even temporarily active) at the same time. They invalidate each other and basically caused a loop of writes. For a long time we wanted to have a better solution for that, since that problem causes compilation to hang with 100% cpu load, which is quite annoying. Also the eventual consistency can cause some unexpected writes to be done when the content or filename are on a temporary value that has not stabilised yet. The problem applies to every side effect in tasks. ### What? To solve that this refactoring changes the way side effects in tasks are handled. Side effects in tasks should not be directly executed anymore. Instead one should emit an "effect" (`effect(async { ... })` helper method) to declare a side effect that should be executed. Top level execution is usually wrapped in a strongly consistent read of the top level task. Here one need to apply the "effects" (`apply_effects(operation).await?` helper method). Since this happens after the strongly consistent read, only effects that have been stabilised are executed. It's also possible to call `apply_effects` somewhere nested in the execution (after a strongly consistent read) if side effects are necessary for further execution. e. g. we do that to emit the files for the node.js pool before executing the pool. Internally effects are collectibles that are propagated through the call graph. Each effect can only be executed once, so updates to the graph will only apply new effects while existing effects are skipped as they were already executed. The actual change is pretty simple, but `FileSystem::write` no longer returns a `Vc<Completion>` but nothing instead. This makes the change needed a bit bigger. Good news: To emit a whole (potential circular) graph we could use recursion now instead of needing to walk the graph while avoiding cycles. This might allow some performance improvements for incremental builds.

Implements support for passing MDX related plugins as strings instead of as JS functions, this makes sure you can use MDX plugins with Turbopack. This still means that e.g. the options have to be serializable, in testing a few plugins these plugins usually have serializable options though. Fixes #71819 Fixes PACK-3345

@ztanner

To prevent this warning when only using Webpack ``` ▲ Next.js 15.0.2 - Experiments (use with caution): · turbo ``` cc @ztanner https://vercel.slack.com/archives/C03S8ED1DKM/p1731604512615979?thread_ts=1731602213.367809&cid=C03S8ED1DKM

### What? Change the option to a boolean, the number version is weird and unexpected. Closes PACK-3418

Works around nodejs/node#52102 Webpack will be harder to patch and I don't know yet where the Webpack runtime is using `vm.runInThisContext`. Best way to observe change is in #71909 and `pnpm test-dev-turbo test/development/app-dir/dynamic-io-dev-errors/ -t "should display error when component accessed data without suspense boundary"`. Before: ``` Error: [...] at Page [Server] (<anonymous>) at InnerLayoutRouter (.next/server/chunks/ssr/[root of the server]__088b8c._.js) ``` After: ``` Error: [...] at Page [Server] (<anonymous>) at InnerLayoutRouter (node_modules/.pnpm/file+..+next-repo-2186278d37ae48e2b1397aa86060054592669ed9909b91e96c5af049d12c04c3+packages+n_vk62popbath6ofshgazcbhzdoe/node_modules/next/dist/src/client/components/layout-router.tsx:324:2) ``` Next step is figuring out why the now sourcemapped frames aren't ignore-listed.

The host header is necessary for determining the current domain locale for ISR so we need to ensure it's in the default allowed headers list. Without this header we aren't able to render links accurately during revalidations. x-ref: #71848 x-ref: NEXT-3842

adds types for `renderOpts.collectedRevalidate` & friends (only used in route modules). nicer than doing `(context.renderOpts as any)` everywhere

Noticed this after rebuilding on top of #71688 Most Linux distributions default to `gcc` for their default `cc` implementation (e.g. via Debian's `build-essential` meta-package, which we can probably implicitly assume is installed), which means that they won't typically have `clang` installed. `librocksdb-sys` seems to always depend on `clang` on Linux. It looks like this might be a consequence of using [rust-bindgen](https://github.com/rust-lang/rust-bindgen) (though I'm confused why this didn't come up earlier), as it looks like rocksdb can otherwise build with gcc. This shouldn't be an issue for macos, since `clang` is the default `cc` on that platform.

…#71692) ### What? We can now enable tree shaking for modules that imports `next/dynamic`. ### Why? The bug is now fixed. ### How? Closes PACK-3231

Trying to fix : #72522 ### What? The finally method actually creates a new promise. You have to listen to this new promise instead of the original one or you might end up with unhandled rejection. Taken from the documentation : > The finally() method of [Promise](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Promise) instances schedules a function to be called when the promise is settled (either fulfilled or rejected). It immediately returns another [Promise](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Promise) object, allowing you to [chain](https://developer.mozilla.org/en-US/docs/Web/JavaScript/Guide/Using_promises#chaining) calls to other promise methods. ### Why? see : #72522 ### How? Fixes : #72522 Co-authored-by: Wyatt Johnson <[email protected]>

### Updating Examples - Remove unnecessary in Dockerfile <img width="398" alt="image" src="https://github.com/user-attachments/assets/63e6df3a-e9b8-42ed-bdd9-dd736131abbe"> Signed-off-by: ductnn <[email protected]> Co-authored-by: JJ Kasper <[email protected]>

Followup to #72241 and #72467 Documenting the new Turbopack support for MDX plugins (remark, rehype, recma plugins): - #71819 (comment)  ### What? Clarify that MDX-related plugins (remark and rehype plugins) are supported when options are serializable, by using the new string name config format ### Why? Support was added in 72802 - #72802 ### How? Add a new section about using plugins with Turbopack

…2870) ## Why? Removing a log that holds very little meaning these days. ![cleanshot_2024-11-03_at_09.13.12_2x.png](https://graphite-user-uploaded-assets-prod.s3.amazonaws.com/FEMDaE9e61ZMCiJM0S0I/ef9c17a7-4fd1-4c70-94c3-c2b31f10c357.png) Co-authored-by: JJ Kasper <[email protected]>

Closes PACK-3368 Closes #72180 Previously, ```js if (sideEffectFalse((state += 1))) { return } ``` was replaced with ```js if ("TURBOPACK compile-time falsy", 0) { return } ``` which discards the side effects of both the arguments, and of the function itself.

This matches the handling added in #70997 but for the Webpack build worker. This unifies the worker code to be the same as the one for the static export worker which already handles this case. When the worker exits ungracefully, we need to propagate an error signal to the parent process so that it doesn't hang. Fixes #67097 Closes NDX-485 Closes #67154

Fixes #72915 Fix the `MONGODB_URI` env variable not found during `npm run build` by moving the variable assignment to `dbConnect` function Co-authored-by: Sam Ko <[email protected]>

### What? small follow-up to fix the tracing of Effects::apply

## Summary Update incorrect code block languages at [redirecting document](https://github.com/vercel/next.js/blob/canary/docs/01-app/02-building-your-application/01-routing/07-redirecting.mdx). ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide Co-authored-by: Delba de Oliveira <[email protected]>

## Summary Update code block language in [error pages](https://github.com/vercel/next.js/tree/canary/errors). ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide Co-authored-by: Delba de Oliveira <[email protected]>

### What? Update the example code with Codemods applied in [Migrate to async Dynamic APIs](https://nextjs.org/docs/app/building-your-application/upgrading/codemods#migrate-to-async-dynamic-apis) docs Co-authored-by: Delba de Oliveira <[email protected]>

### What? Fix broken link to the `rootDir` doc at https://nextjs.org/docs/messages/no-html-link-for-pages#options ### Why? This is because there is an extra space between a path and an anchor. ### How? This PR removes it to fix the format. Co-authored-by: Delba de Oliveira <[email protected]>

## Description Follow up #52038. Remove `https://nextjs.org` from URL in [error pages](https://github.com/vercel/next.js/tree/canary/errors). ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide --------- Co-authored-by: Delba de Oliveira <[email protected]>

## Summary Add the paragraph name to [next-prerender-crypto](https://github.com/vercel/next.js/blob/canary/errors/next-prerender-crypto.mdx) error page like [other next-prerender pages](https://github.com/vercel/next.js/blob/canary/errors/next-prerender-current-time.mdx). ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide Co-authored-by: Delba de Oliveira <[email protected]>

### What? Simplify Vapid key generation docs ### Why? Creating a script is tedious. The CLI can be used across projects without needing to reference and recreate this script. ### How? Added instructions to install `web-push` globally and how to run it. --------- Co-authored-by: Delba de Oliveira <[email protected]>

## Description Unify the note format according to the [contribution guide](https://nextjs.org/docs/community/contribution-guide#notes). In addition, add missing bullet points to useful links. ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide --------- Co-authored-by: Delba de Oliveira <[email protected]>

#71241) ### What? Implement side effect optimization without creating a new `ModulePart`. This PR is `partial` because this PR does not handle the exclusion of a non-existent module perfectly. #72947 will do it. ### Why? #70336 (comment)

…pdate the demo (#72644)  ### What? Removed the deprecated `iframe` attribute `frameborder`, adjusted `allowfullscreen` to be `allowFullScreen` and updated the demo video. ### Why? - `frameborder` is marked as deprecated ([MDN](https://developer.mozilla.org/en-US/docs/Web/HTML/Element/iframe#frameborder), [DefinitelyTyped](https://github.com/DefinitelyTyped/DefinitelyTyped/blob/master/types/react/index.d.ts#L3231-L3232)) - `allowfullscreen` should be `allowFullScreen` ([DefinitelyTyped](https://github.com/DefinitelyTyped/DefinitelyTyped/blob/a5dd14a91255869c8a929284c2531bd0e96d28ff/types/react/index.d.ts#L3229)) - The YouTube `iframe` demo has an invalid embedded URL, fixed and updated to the latest Next.js Conf as well. --------- Co-authored-by: Delba de Oliveira <[email protected]>

## Summary Change `jsx` to `diff` in [next-dynamic-api-wrong-context](https://github.com/vercel/next.js/blob/canary/errors/next-dynamic-api-wrong-context.mdx) error page. ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide Co-authored-by: Delba de Oliveira <[email protected]>

What is the documentation issue? There is incorrect information about how fetch works in the new Next.js version (^15.0.0). The documentation says that fetch is automatically cached in the said version but now by default fetchs are no longer being cached Is there any context that might help us understand? [Link of the section](https://nextjs.org/docs/app/building-your-application/caching#fetch-optionscache) Does the docs page already exist? Please link to it. https://nextjs.org/docs/app/building-your-application/caching#fetch-optionscache --------- Co-authored-by: Delba de Oliveira <[email protected]> Co-authored-by: Sam Ko <[email protected]>

#72847 missed this

The bread and butter of NFT. The next two PRs split the work into two halves, one half plumbs as far as ResolveResultItem, and the next one fills in the remaining (and fixes some bugs). Closes PACK-3379

- Change the Next manifest output assets to also track references to output assets there. That way, `page.js.nft.json` lists the referenced output assets of `page.js` and e.g. `page_client-reference-manifest.js` and all the chunks references in the manifest as well. - For `ResolveResultItem::External`, continue resolving with a fresh `ModuleAssetContext` that doesn't have any of the build-time settings (because the Node environment when running the server won't have these settings either). Closes PACK-3380

)

Closes PACK-3421 Address PR feedback from #72305 When working on this, I uncovered another bug which will be handled in a different PR. Changes: - Don't rely on `DiskFileSystem` but only via file paths and `is_inside()` - Put the `RebasedAsset` for externals on the `[project]/` fs (this way, they aren't emitted either), i.e. don't perform any rebasing at all and just reference the files where they are --------- Co-authored-by: Niklas Mischkulnig <[email protected]>

…Vc (#72889) These changes are generated using https://github.com/vercel/turbopack-resolved-vc-codemod, using the fix to the logic that finds `Vc` field: vercel/turbopack-resolved-vc-codemod@5c277f8 This is a combination of applying compiler hints and an LLM to fix some easy cases.

## Fixing a bug ### Related issues fixes #64031 ### Why is this change needed Per the Next.js [docs](https://nextjs.org/docs/pages/api-reference/next-config-js/output) > Next.js' production server is also traced for its needed files and output at .next/next-server.js.nft.json which can be leveraged in production. When I tried doing this my [custom server](https://nextjs.org/docs/pages/building-your-application/configuring/custom-server) would crash on startup because of this codepath https://github.com/vercel/next.js/blob/canary/packages/next/src/server/config.ts#L1022 My first attempt at fixing this involved changing ``` if (!process.env.__NEXT_PRIVATE_STANDALONE_CONFIG) { ``` to ``` if (phase === PHASE_PRODUCTION_SERVER) { ``` but this is not guaranteed to work because the next.config.js and it's required dependencies may not be present causing the server to use the default config The best solution is to mimic standalone mode as closely as possible ### What does this change do This change allows a custom server running in production with only required files to pass in a preloaded next config (most likely from `required-server-files.json`) This is very similar to how the [standalone server works](https://github.com/vercel/next.js/blob/bc49287063d6afd50a7361566b14f1ab6c26b136/packages/next/src/build/utils.ts#L2142) ### Testing I'd like to get an initial round of feedback from the next team before I dive into writing an integration test. I patched nextjs in my repo and this solution worked

By adding `visit_mut_function` we can deduplicate the code that's common between `visit_mut_fn_expr` and `visit_mut_fn_decl`. And more importantly, this prepares us for handling the transform of method props, which will also use `visit_mut_function`. As a positive side effect this brings two small improvements with it: - always highlight the whole function when `async` is missing, not only the identifier - assign a function name to an anonymous arrow function expression in a variable declaration

With this PR, we are now allowing `"use server"` and `"use cache"` to be used in method properties, e.g.: ```js export const obj = { async foo() { 'use cache' return Math.random() }, async bar() { 'use server' console.log(42) }, } ``` Or, more realistically, something like this: ```js export const api = { product: { async fetch() { 'use cache' return fetch('https://example.com').then((res) => res.json()) }, }, } ``` Supporting this pattern is mostly done for convenience; it allows users to group server functions into a common object, using method properties. Notably, we will not support the usage of `this` and `super`, for which we will add build errors in a follow-up PR.

We now look through each section for the source and use that section to determine if a frame is ignore-listed. Previously, stacks from Turbopack sourcemaps were not ignore-listed since they were index maps.

I commonly add a `debug` script that allows me to attach a debugger when running Next.js in the isolated test dirs. Now we always add it and make it work just like `debug` from the monorepo. So after the test you can `cd` into the created test directory and run `pnpm debug dev`. `debug` is just an alias for `next` with some additional configurations: - tracing deprecations - no telemetry - no canary checks - debugger attachable - source maps `debug` also ensures the SWC version that was used during the test is used. Previously it used the cached version which isn't necessarily the one built in this specific worktree.

- `__turbopack_external_require__("node:buffer")` becomes `__turbopack_external_require__("node:buffer", () => require("node:buffer"))`. - This way, running esbuild on the Turbopack output will correctly detect usage of Node builtins and is able to polyfill them - To make that `require` the global (non-shadowed) one, change the previous `t: require` item from `__turbopack_context__` to `t: __turbopack_require_real__` and the fake require added in #70255 to `__turbopack_stub_require_` Closes PACK-3389 --------- Co-authored-by: Niklas Mischkulnig <[email protected]>

No significant changes to runtime detected.

Co-authored-by: Jiachi Liu <[email protected]>

Reverts #72966

In the Segment Cache implementation of prefetching, during a navigation, it's possible that the prefetch for a segment may already be in-progress, but we haven't received its data yet. In that case, since the prefetch is very likely to load faster than the dynamic request, we might as well use the result, if we can. This updates type CacheNode.loading to accept a promise for the loading module data, and updates the LayoutRouter to conditionally unwrap it if needed. We already use this same strategy in other places, like for streaming in the main segment data.

When generating the segment responses, we should use the `prerender` API instead of the `renderToReadableStream` API to ensure that when we abort the stream, dynamic access is encoded as a hanging reference, rather than as an error. I didn't notice this was wrong until I managed to get an initial version of the client Segment Cache running locally and the errors were surfaced in the app. No tests yet since it's not really possible to test until the client is implemented, which I'm working on now. But you can look at the build output and notice that the errors are no longer present.

Based on: - #72874 --- The <head> does not belong to any particular segment; it represents the entire page. Except in the case where two URLs rewrite to the same result, it's very unlikely to benefit from deduplication, so there's no benefit to caching it separately from the route tree. So we'll send it in the same response. Since the head may contain dynamic data, the route tree response may now contain hanging promises, so we'll need to use the same AbortController trick that we use for the segments.

The `<Image>` component doesn't optimize SVG images since they are vector images and can already scale up and down to any width. Therefore, the component will try to detect SVG images based on the file name to avoid sending them to the Image Optimization API at all. This is because the default behavior of the Image Optimization API is to error for SVG images due to the security implications of `<svg>` with nested `<script>`, etc (unless `dangerouslyAllowSVG` is enabled. However, we don't want users to reach for `dangerouslyAllowSVG` because it is not usually what they intended. So instead, we can improve the heuristic for detecting SVG in the `<Image>` component by ignoring the query string. Related to nodejs/nodejs.org#7244

Based on: - #72874 - #72890 - #72872 --- This adds an initial implementation of the client Segment Cache, behind the experimental `clientSegmentCache` flag. (Note: It is not anywhere close to being ready for production use. It will take a while for it to reach parity with the existing implementation.) I've discussed the motivation in previous PRs, but I'll share a brief summary here again: The client Segment Cache is a rewrite of App Router's client caching implementation, designed with PPR and "use cache" in mind. Its main distinguishing feature from the current implementation is that it fetches/caches/expires data per route segment, rather than per full URL. An example of what this means in practical terms is that shared layouts are deduplicated in the cache, resulting in less bandwidth. There are other benefits we have in mind but that's the starting point. I've tried to extract the work here into reasonably-sized commits (many of which have already landed) but this one here is sorta unavoidably large. Here are the main pieces: - [segment-cache/cache.ts](https://github.com/acdlite/next.js/blob/initial-implementation-client-segment-cache/packages/next/src/client/components/segment-cache/cache.ts): This module is where the cache entries are maintained in memory. An important design principle is that you must be able to read from the cache synchronously without awaiting any promises. We avoid the use of async/await wherever possible; instead, async tasks write their results directly into the cache. This also helps to avoid race conditions. Currently there's no eviction policy other than stale time, but eventually we'll use an LRU for memory management. - [segment-cache/scheduler.ts](https://github.com/acdlite/next.js/blob/initial-implementation-client-segment-cache/packages/next/src/client/components/segment-cache/scheduler.ts): This module is primarily a task scheduler. It's also used to manage network bandwidth. The design is inspired by React Suspense and Rust Futures — tasks are pull-based, not push-based. The backing data structure is a MinHeap/PriorityQueue, to support efficient reprioritization of tasks. - [segment-cache/navigation.ts](https://github.com/acdlite/next.js/blob/initial-implementation-client-segment-cache/packages/next/src/client/components/segment-cache/navigation.ts): This module is responsible for creating a snapshot of the cache at the time of a navigation. Right now it's mostly a bunch of glue code to interop with the data structures used by the rest of the App Router, like CacheNodeSeedData and FlightRouterState. The long term plan is to move everything to using the Segment Cache and refactoring those data structures. Additional explanations are provided inline.

The edge runtime does `await self._ENTRIES["middleware_..."].default.call(...)`, so instead export a Proxy that hides that the module object itself is a promise. Partial revert of #62141

## Why? There's a typo in the note—Route Handler should be singular.

…` helper function (#73016) Dead code strikes again. https://github.com/est31/warnalyzer has been updated for recent rust toolchains, so I might give that a shot soon to see if I can clean up all of our dead functions.

### What? Allow KeyValueDatabase to implement to variants of WriteBatches * Serial: It's not Send+Sync and writing takes a `&mut self` * Concurrent: It's Send+Sync and writing takes a `&self` When using the concurrent variant we can avoid buffering the intermediate results and write that directly to the write batch. But the database need to take care of handling concurrency.

### What? * make invalidation more granular * fix untagged enum serialization * fixes ppr config option serialization * fix typo * improve tracing * less aggressive persisting * fix verify_serialization feature

### Why? This PR added an indication of where the env was loaded from when `experimental.typedEnv` was enabled. Also, it allows the user to set `NODE_ENV=production` to enable the `typedEnv` feature for `.env.production*` files. ![CleanShot 2024-11-21 at 19 48 20](https://github.com/user-attachments/assets/ce3c7180-f26a-4378-a74f-a5998a363211) ### How? Modified `@next/env` to pass parsed envs along with the `loadedEnvFiles` value and used the location to indicate via JSDoc.

### What? Enable single-incoming-edge optimization for module fragments with `ModulePart::Export`. ### Why? It will reduce the number of modules greatly. ### How? Closes PACK-3481

There's significant drift between how Link is implemented in pages vs app dir, and we're planning to make even more changes. (For example, we want to change the intersection observer prefetch behavior, but only in app router) To simplify this drift in behavior, this will alias `next/link` to `client/app-dir/link.tsx` in app dir bundles. There is no change in behavior to the `link` component in this PR. The next PR in the stack will remove the current forked handling to remove things that are irrelevant to the router it's rendered in.

Adds a separate bench for Turbopack build enabled.

This follows up the work from #73019 to update the implementations for the forked Link components to remove Pages specific handling in the App link, and App specific handling in the Pages link.

This is meant to add a bit more clarity of why we can't set cookies to the browser in a server component even if it's through a server action or route handler. Co-authored-by: Delba de Oliveira <[email protected]>

This pull request adds the Next.js template for Fleek. Users can use it to get started using full-stack Next.js on Fleek. This template is directly from the official Fleek community templates' collection. ### What? This adds to the templates references within the docs to get Fleek on the Next.js documentation. ### Why? Community templates of Next.js are great to help spread the word that Next.js can be deployed anywhere.  Co-authored-by: Delba de Oliveira <[email protected]>

## Why? The default fetch behavior, which is shown as `auto no cache` when you set the config [logging](https://nextjs.org/docs/app/api-reference/next-config-js/logging) to `true`, essentially means fetch once on `next build` to prerender the page. But, when a [Dynamic API](https://nextjs.org/docs/app/building-your-application/rendering/server-components#dynamic-rendering) is used, it will fetch from the resource on every request. This is different from `{ cache: 'no-store' }`, which regardless of whether Dynamic APIs are used, will de-op the page from prerendering.

Fix mdx formatting. Previously MDX was not formatted correctly. (see image - `prefetch=`) https://nextjs.org/docs/app/api-reference/next-config-js/staleTimes <img width="659" alt="Screenshot 2024-11-19 at 15 11 57" src="https://github.com/user-attachments/assets/44f77fd4-657e-409a-9190-9430fd96d5f8"> --------- Co-authored-by: Delba de Oliveira <[email protected]>

Hi, Team. The path of route handler seems to be missing `api` and I'd like to fix it. (`middleware.ts` above the route handler file, it requests to `/api/redirects..`) <img width="650" alt="image" src="https://github.com/user-attachments/assets/f9ff3034-de36-42ec-8d85-a4a923005fdf"> Thank you :) ### Improving Documentation - correct route handler path --------- Co-authored-by: hyungji <[email protected]> Co-authored-by: Delba de Oliveira <[email protected]>

This prop is a no-op in app router. This removes the code that handled it for the App Router link, and adds a warning for apps that are attempting to use it in an app router context.

Reverts #72988 Right now, we're focused on making sure the examples linked are guaranteed to work and placed inside the Next.js community GitHub organization.

### What? The bench need to run in the correct directory

When handling a redirect through a server action, we pass in the current href rather than the destination href to `handleExternalUrl`, which is incorrect since that won't reflect the updated URL from a server action's redirect response. While looking into this, I noticed a handful of server action branches that weren't properly rejecting or resolving the action promises. This would cause the transitions to stall until they were fulfilled, so I updated all of the existing branches where we early return to also resolve the promise. Fixes #72842

- The waitFor function now supports a predicate. Unlike retry, waitFor does not include a timeout. Use it only when the predicate is guaranteed to evaluate to true before proceeding to the next test. If it runs indefinitely, Jest enforces a 120-second timeout per test. - The getCliOutputFromHere utility provides a convenient string getter that captures CLI output in between.

## Why? Add documentation for experimental `inlineCss`. ``` import type { NextConfig } from 'next' const nextConfig: NextConfig = { experimental: { inlineCss: true, }, } ``` - x-ref: #72195

## Why? The `latest` tag currently points to v15, so we need to update the tag to `next-14`.

…ding (#73030) ## History In a previous PR #65846 the chroma subsampling setting was removed for AVIF images. This caused the default of `4:4:4` to be used which increased the image size and CPU a lot. I considered switching back to `4:2:0` but found that reducing the `effort: 3` actually yielded significantly better performance with only a slight difference in size. ## Benchmark For example, take this PNG that has a width of 1805. ``` curl -JO https://images.bookroo.com/site/landing/homepage/book-collection.png ``` Using `w=1850` and `q=75`: ``` Current: 2_778ms and 153_364 bytes Chroma: 2_214ms and 132_927 bytes This PR: 1_265ms and 158_862 bytes ``` Using `w=1500` and `q=75`: ``` Current: 625ms and 95_810 bytes Chroma: 524ms and 89_500 bytes This PR: 185ms and 99_127 bytes ``` Using `w=1200` and `q=75`: ``` Current: 506ms and 74_738 bytes Chroma: 407ms and 68_998 bytes This PR: 144ms and 77_955 bytes ``` For this particular image, the size will increase a bit but there are other images I tested where the size was actually smaller. So its a big improve in CPU usage compared to reverting to the chroma subsampling setting. We also are keeping AVIF smaller than WebP which is the expected behavior. ## Learn More [Chroma subsampling](https://en.wikipedia.org/wiki/Chroma_subsampling) is the practice of encoding images by implementing less resolution for [chroma](https://en.wikipedia.org/wiki/Chrominance) [information](https://en.wikipedia.org/wiki/Information) than for [luma](https://en.wikipedia.org/wiki/Luma_(video)) information, taking advantage of the human visual system's lower acuity for color differences than for luminance Image compression deep-dive video: https://www.youtube.com/watch?v=F1kYBnY6mwg Sharp Options: https://sharp.pixelplumbing.com/api-output#avif

# Context [Test run](https://github.com/vercel/next.js/actions/runs/11962411287/job/33382985253?pr=73030#step:29:421) fails due to ``` ▲ Next.js 15.0.4-canary.22 - Local: http://[::1]:35903 - Network: http://[::]:35903 ✓ Starting... Error: "next start" does not work with "output: export" configuration. Use "npx serve@latest out" instead. at <unknown> (dist/server/next.js:215:31) at async NextServer.prepare (dist/server/next.js:154:24) at async initializeImpl (dist/server/lib/render-server.js:92:5) at async initialize (dist/server/lib/router-server.js:436:22) at async Server.<anonymous> (dist/server/lib/start-server.js:272:36) ``` However, there was no `output: export` specified in the [test setup](https://github.com/vercel/next.js/tree/jude/which-config-are-we-using/test/integration/edge-runtime-configurable-guards). **Mysterious!** # Hypothesis `next start` recursively finds the `next.config` upwards in the directory tree and happens to locate an extraneous `next.config` that specified `output: export`. # Changes - Output the config file path in test mode for future debugging - Short-circuit `findUp` for the broken test by placing an empty config in the project directory

x-ref: #72634 (comment)

## Improving Documentation dynamic `params` from `generateStaticParams` is now a promise but is now promise in type. Type has been changed for every Promise `params`. Co-authored-by: Delba de Oliveira <[email protected]> Co-authored-by: JJ Kasper <[email protected]>

Hi, Team. It seems Promise is missing when defining params type. --------- Co-authored-by: hyungji <[email protected]> Co-authored-by: Maikel <[email protected]>

…on) (#73018) This PR has two main goals: - **Consolidate the detection of server directives in modules and function bodies.** Previously, these were two separate implementations with significant overlap, but also some [questionable discrepancies](#72811 (comment)). - **Model the current directive (in a file or function) using an enum instead of two separate booleans.** This change prevents any confusion that both directives might be present simultaneously. Additionally, we're now emitting an error if multiple directives (`"use server"` and `"use cache"`) are defined in the same location (at the top of a file or function body). A mixed usage of `"use server"` and `"use cache"` _across different locations_ is still allowed, e.g. `"use cache"` at the top of a file, and `"use server"` in a function. > [!NOTE] > The diff may be tricky to review because chunks from two different functions are combined into a single function. Fortunately, we have comprehensive test coverage for the transforms, which instills high confidence that these changes are correct.

- `expirePath` and `expireTag` are not yet available in a stable version, so we don't want these APIs leaking into stable docs.

## Improving Documentation I just add **Good to know** to below path note which was in a block but not add the tag. > Pages > Rendering > Automatic Static Optimization **Before:** > Statically generated pages are still reactive: Next.js will hydrate your application client-side to give it full interactivity. **Now:** > **Good to know:** Statically generated pages are still reactive. Next.js will hydrate your application client-side to give it full interactivity. Co-authored-by: Sam Ko <[email protected]>

Co-authored-by: Sam Ko <[email protected]>

…3053) While the React Compiler is still in beta, this functionality was not yet documented. I also updated this to remove the 15 RC since 15 is now stable. Co-authored-by: Sam Ko <[email protected]>

### What This PR changes `NextCustomServer` to no longer subclass `NextServer` and instead just delegate all methods to `this.renderServer`. I've also added an interface implemented by both `NextServer` and `NextCustomServer` to make it explicit what we're actually exposing. (Note that we're currently exposing too much for legacy reasons. we might want to decide to clean this up a bit, i've started that in #73021) I've also cleaned up some random `(_ as any)`s and unnecessary indirections around the area. ### Why currently, `NextCustomServer` is a subclass of `NextServer`, which makes it very confusing to reason about, because in `prepare()` it creates *a nested `NextServer`* (stored in `this.renderServer`). this isn't immediately visible in the code, but that's what `getRequestHandlers` ends up doing. https://github.com/vercel/next.js/blob/58b993c239689f22641682bbcb0d8dea6d85fdf1/packages/next/src/server/next.ts#L281 then, it delegates *some* methods to that (`render`, `getRequestHandler`, `setAssetPrefix`). But some methods (e.g. `render404`) *aren't* overridden/delegated and just run the implementation inherited from `NextServer`. And those implementations are unaware of `this.renderServer`, so they end up initializing a *second* `NextNodeServer` (`this.server` vs `this.renderServer.server`) and run against that instead. that's bad! and it's why e.g. #61676 was necessary -- we needed to update both servers! i think that this is a leftover of the old implementation where we ran a separate process for rendering. for historical context, see: - #49805 which seems to have introduced this split (the proxy only overrides `prepare`, `getRequestHandler`, and `render`) - #53523 which turned the proxy code into `NextCustomServer` also apparently `render404` and others were broken in development (try running the tests in the first commit I've changed `NextCustomServer` and see the server hang forever).

## Description Follow up #72617. I have already fixed at #72944. `app/api/redirects/route.ts` is ts extension. ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide Co-authored-by: Sam Ko <[email protected]>

## Why? Update params type to be async (v15). - x-ref: #72070

## Why? The default config of fetch, which is `auto no cache`, is slightly different from the fetch option `{ cache: 'no-store' }` (this opts the whole route out of statically prerendering), so we need to clarify that further. - https://nextjs.org/docs/canary/app/api-reference/functions/fetch#optionscache - x-ref: #72982 - Fixes #73109

…73058) This is a follow-up fix for #72969. Method or function properties of exported objects in `"use cache"` or `"use server"` files must not be compiled into server functions, unless they have an inline server directive.

Accessing `this` or `arguments` in server functions is not allowed. With this PR, we are now emitting build errors in this case. --------- Co-authored-by: Janka Uryga <[email protected]>

…73106)

#73136)

…" (#73138) This relands #72921 to be a more minimal refactor. This re-applies the core change to use the existing worker that has proper error bubbling handling without changing the worker lifecycle. For the purposes of the internal OOM we were seeing, this ensures that any custom `max-old-space-size` flags are preserved during the webpack build step, even when using the shared worker. Instead, I moved it to `createStaticWorker`, as that was where it was intended to be respected when it landed in #46705.

Co-authored-by: vercel-release-bot <[email protected]>

…2936)

### What? Update `lightningcss` to the latest. - https://github.com/parcel-bundler/lightningcss/releases/tag/v1.28.2 ### Why? It's released.

- adds per-module tracing for `find server actions` (it's implemented in the same way for next/dynamic and client references). Now, `find server actions` isn't just a single blob anymore. - add a tracing span for `all_server_paths`

Turbopack has supported instrumentation.js for a long time now

When a `"use cache"` module has no exported or annotated cache function, we can omit the import statement for the cache wrapper function.

With this PR, we are allowing _static_ class methods to be annotated with `"use cache"` or `"use server"`. Class _instance_ methods on the other hand are not allowed as server functions.

**Do not merge until #72785 lands in canary.** Closes: - https://linear.app/vercel/issue/DOC-3820/[unstable]-forbidden-and-forbiddenjs - https://linear.app/vercel/issue/DOC-3849/[unstable]-unauthorized-and-unauthorizedjs - https://linear.app/vercel/issue/DOC-3860/[experimental]-authinterrupts-config-option

We forgot to update the manifest when merging the NFT PRs

### Why? As we upgraded create-next-app ESLint to v9, use flat config (with compatible function) for the templates. `eslint.config.mjs`: ```mjs import { dirname } from "path"; import { fileURLToPath } from "url"; import { FlatCompat } from "@eslint/eslintrc"; const __filename = fileURLToPath(import.meta.url); const __dirname = dirname(__filename); const compat = new FlatCompat({ baseDirectory: __dirname, }); const eslintConfig = [...compat.extends("next/core-web-vitals")]; export default eslintConfig; ``` As `import.meta.dirname` was introduced in Node.js version v20.11.0, we can simplify to when on minimum version: ```mjs import { FlatCompat } from "@eslint/eslintrc"; const compat = new FlatCompat({ baseDirectory: import.meta.dirname, }); const eslintConfig = [...compat.extends("next/core-web-vitals")]; export default eslintConfig; ``` But best is that other dependency plugins support flat config so we can upgrade to pure Flat Config as well. Closes NDX-374

Depends on #73010 A very small speedup, mostly noticeable in cpu time, not really in wall time ``` before: 504.86s user 73.31s system 792% cpu 1:12.93 total 522.09s user 76.70s system 778% cpu 1:16.94 total 503.73s user 68.78s system 818% cpu 1:09.92 total after: 471.20s user 61.73s system 757% cpu 1:10.33 total 481.69s user 69.05s system 805% cpu 1:08.33 total ```

### What? Use `ResolvedVc<T>` instead of `Vc<T>` for struct fields in `turbopack-css` ### Why? ### How?

See #73105 (comment)

This is extracted from #72513. This override appears to break the build for rocksdb. It doesn't seem like there's any good reason for us to do this. It looks like we used to do a cross-build for macos arm64 (#28138), which would explain the need to override the sysroot, but we don't do that anymore. We don't do this override for x86-64 builds (even though that one _is_ a cross-build? The napi cli probably does the right thing for us...). Manually triggered `build-and-deploy` CI job: https://github.com/vercel/next.js/actions/runs/12018322261

### Adding or Updating Examples Hi, Team. Updating example to await params as it is now used asynchronously. Co-authored-by: JJ Kasper <[email protected]>

) Hi, Team. Updated Reusing data across multiple functions examples on data fetching documentation to improve type safety and accuracy. ### Summary of Changes 1. Added Async Params 2. Fixed Dynamic Route Example Path - Corrected the path structure for dynamic routes to reflect actual usage as `app/page.tsx` cannot be used to access dynamic routed paths. 3. Improved Type Safety in `generateStaticParams` - Type mismatch: `generateStaticParams` expects string parameter, but examples used `post.id` which of type is number, causing both build time and client run time failures. <img width="1109" alt="image" src="https://github.com/user-attachments/assets/ae6fad44-0b7f-4aac-ac31-74a6581c995b"> <img width="714" alt="image" src="https://github.com/user-attachments/assets/d8fb19c9-9dd1-4e6a-8d2b-b25f719e3742"> 4. Replaced let variable with const variable - Prevent accidental reassignment as it is not needed --------- Co-authored-by: JJ Kasper <[email protected]>

…s, not remote images (#73184) Currently, enabling skew protection will cause new deployments to invalidate the browser cache since the deployment id is including in the query string as `dpl`. In many cases, such as JS and CSS, these assets are bundled and include a content hash in the file name, so new deployments already invalidate browser cache. However, the optimized image cache is preserved between deploys, so enabling skew protection breaks that expectation. This PR removes the `dpl` query string for remote images since those files don't live in the deployment and therefore don't need skew protection. - Related to #73015

This PR replaced eslint-related docs code examples from `.eslintrc.json` with `eslint.config.mjs`. Closes NDX-373

### What? Use `ResolvedVc<T>` instead of `Vc<T>` for struct fields in `turbopack-core` ### Why? ### How?

… `turbopack-wasm` (#73196) ### What? Use `ResolvedVc<T>` instead of `Vc<T>` for struct fields in `turbopack`, `turbopack-tests`, `turbopack-wasm`. ### Why? ### How?

### What? Use `ResolvedVc<T>` instead of `Vc<T>` for struct fields in `turbopack-nodejs`. ### Why? ### How?

### What? Use `ResolvedVc<T>` instead of `Vc<T>` for struct fields in `turbopack-env`. ### Why? ### How?

Closes: https://linear.app/vercel/issue/DOC-3696/page-layouts-and-pages

Going through the docs with fresh eyes, picked up some nits.

aka "one graph" - Make client references actual references (non-chunkable, so chunking ignores them right now. We can change that later and move more chunking into Turbopack itself as opposed to the manual stuff in `AppEndpoint`) - `find server actions` now just walks the rsc_entry and discovery client references itself - `collecting next/dynamic` still walks the RSC graph which is wrong (as before). I left this to make this PR a noop that shouldn't change any behavior. We'll fix this in the coming reimplementation (and add tests for what is apparently currently not asserted 😐) **View without whitespace changes** Closes PACK-3532

Closes: https://linear.app/vercel/issue/DOC-3886/[page]-turbopack - Moves `Turbopack` page under API Reference - Shares page between `app` and `pages` docs - Updates page to follow template Redirects: vercel/front#38704

Changed example to useActionState  The current docs for auth using useFormState that throws error as renamed , hence changed the example to use useActionState --------- Co-authored-by: Delba de Oliveira <[email protected]>

This allows us to skip unnecessary work, and also avoids generating invalid code, e.g. not allowed `super` calls in hoisted methods.

#73061)

@samcx

#70567) This PR updates the with-styled-jsx-scss example for using the App Router. Here are the changes that have been made: - Renamed the "pages" folder to the "app" folder. - Added the layout.tsx file as part of the App Router. - Updated the package.json file. CC: @samcx --------- Co-authored-by: Sam Ko <[email protected]>

@samcx

…ter (#70663) This PR updates the with-styled-jsx-plugins example for using the App Router. Here are the changes that have been made: - I renamed the "pages" folder and moved it to the "app" folder. - Added the layout.TSX file as part of the App Router. - Updated the package.json file. CC: @samcx --------- Co-authored-by: qed42 <[email protected]> Co-authored-by: Sam Ko <[email protected]>

@samcx

This PR updates the with-unstated example for using the App Router. Here are the changes that have been made: - Renamed the "pages" folder to the "app" folder. - Added the layout.tsx file as part of the App Router. - Moved the components folder inside app folder. - Updated the package.json file. - Added hooks for counter and clock. CC: @samcx --------- Co-authored-by: samcx <[email protected]>

) This removes a check that appears to be redundant: we only want to warn about the fallback behavior when deployed in an environment where the env vars needed to detect proper fallback values (eg, not deployed to Vercel) aren't available. Since this check is already captured, we don't need both. The rationale for explicitly logging in standalone mode (even in dev) is that `standalone` mode won't ever be deployed to Vercel. But it's equally possible for you to build a Next.js app and not deploy it to Vercel. At best, this was reducing some additional console noise. But at worst, you're deploying an app that will serve a `localhost` URL in production. If you're using a relative URL without a metadataBase, it seems worth explicitly calling out as early as possible that we're going to fill in a fallback for you. Ultimately we should surface this information regardless of env in devtools. But since warning is the default, in #73066 I've updated it to only warn in dev when explicitly passing a relative og URL, rather than always making it warn when using the `opengraph-image` convention.

This refactors some unnecessary indirection that made it trickier for me to understand the flow of ogImage resolution. Specifically: - `getSocialImageFallbackMetadataBase` was taking in `metadataBase` and then inverting it and returning `isMetadataBaseMissing` which is circular. It was also being called higher up in the stack than it needed to be resulting in args being passed through intermediary functions that didn't need it - `validateResolvedImageUrl` was removed in favor of grabbing the fallback and warning appropriately in `resolveAndValidateImage`. This makes it more obvious that when we're warning about using the fallback, we should actually use the fallback, which I've done in #73066

…ages (#73066) When resolving opengraph / twitter images, we currently ignore the configured `metadataBase` value, which goes against the [documentation](https://nextjs.org/docs/app/api-reference/functions/generate-metadata#default-value) and doesn't seem correct, since an explicit `metadataBase` should be preferred over a fallback. This updates the handling to ensure an explicit `metadataBase` takes priority over the implicit path discovery. The one remaining special case here is for the "static metadata" case, aka the `opengraph-image` convention. We'll continue to prefer the fallback `metadataBase` in preview Vercel deploys & locally. Fixes #66957 Closes NDX-260

### What? Use `ResolvedVc<T>` instead of `Vc<T>` for struct fields in `turbopack-dev-server` ### Why? ### How?

### What? Use `ResolvedVc<T>` instead of `Vc<T>` for struct fields in `turbopack-tasks-fetch`. ### Why? ### How?

@bgw

Does what it says on the tin. One struct seems to be trivially used in a trait which requires it to be a Vc so I left it with a comment. @bgw perhaps we can set up some lint + disabling comment for all the cases in this doc? Would be handy to ensure we are continuing to follow the heuristic in the future also.

this PR changes `after`-related code to unconditionally set everything up even if `experimental.after` is disabled, because we want to be able to re-use some of the `after` machinery for instrumentation-related things.

### What? Use `ResolvedVc<T>` instead of `Vc<T>` for struct fields in `next-api`, but split into multiple PRs. ### Why? To reduce scope of changes of each PRs so I can debug HMR failures. ### How?

Allows using `unstable_after` in `generateStaticParams`. definitely niche, but since we're now allowing `after` to be used during build, i see no reason not to. Also fixes an issue where we didn't pass `experimental.after` into the context used for `generateStaticParams`, so if `after` was called, it'd always say that you have to enable `experimental.after` even if you already did.

undici 7 was released 30 minutes ago and broke the tests because Turbopack now has troubles with its `node:sqlite` import. You can see the test failure in #73261 (comment) ``` Error: Cannot find module 'node:sqlite': Unsupported external type Url for commonjs reference at <unknown> (.next/server/chunks/ssr/c1118_undici_e1ac85._.js:35:29147) at [project]/node_modules/.pnpm/[email protected]/node_modules/undici/lib/cache/sqlite-cache-store.js [app-rsc] (ecmascript) (.next/server/chunks/ssr/c1118_undici_e1ac85._.js:35:29278) at instantiateModule (.next/server/chunks/ssr/[turbopack]_runtime.js:590:23) at getOrInstantiateModuleFromParent (.next/server/chunks/ssr/[turbopack]_runtime.js:645:12) at commonJsRequire (.next/server/chunks/ssr/[turbopack]_runtime.js:147:20) at [project]/node_modules/.pnpm/[email protected]/node_modules/undici/index.js [app-rsc] (ecmascript) (.next/server/chunks/ssr/c1118_undici_e1ac85._.js:116:5860) at instantiateModule (.next/server/chunks/ssr/[turbopack]_runtime.js:590:23) at getOrInstantiateModuleFromParent (.next/server/chunks/ssr/[turbopack]_runtime.js:645:12) at esmImport (.next/server/chunks/ssr/[turbopack]_runtime.js:132:20) at [project]/app/undici/page.js [app-rsc] (ecmascript) <internal part 2> (.next/server/chunks/ssr/[root of the server]__83c897._.js:1:5666) { ``` https://github.com/nodejs/undici/blob/1cfe0949053aac6267f11b919cee9315a27f1fd6/index.js#L52-L59 https://github.com/nodejs/undici/blob/1cfe0949053aac6267f11b919cee9315a27f1fd6/lib/cache/sqlite-cache-store.js#L3

These were the only 2 occurrences in the codebase. Might be a candidate for an ast-grep lint rule

As discussed we are removing the deprecated field from `revalidateTag` and `revalidatePath` as they aren't going away anytime soon and the new `expire` APIs will have slightly different semantics so not a 1-1 rename with no arguments change anymore. We are also adding the `unstable_` prefix for `expireTag`/`expirePath` while we iterate on them so it's clear they will change. --------- Co-authored-by: Zack Tanner <[email protected]>

This PR adds the missing `FormData` type annotation to the formData parameter in the server actions example. Co-authored-by: Delba de Oliveira <[email protected]>

### What? Use `ResolvedVc<T>` instead of `Vc<T>` for struct fields in `next-api`, but split into multiple PRs. ### Why? To reduce scope of changes of each PRs so I can debug HMR failures. ### How?

## Description Now that next-tweet has been renamed [react-tweet](https://github.com/vercel/react-tweet), [Static Tweet (Demo)](https://static-tweet.vercel.app/) navigates https://react-tweet.vercel.app. So updated the link. ### Adding or Updating Examples - [x] The "examples guidelines" are followed from our contributing doc https://github.com/vercel/next.js/blob/canary/contributing/examples/adding-examples.md - [x] Make sure the linting passes by running `pnpm build && pnpm lint`. See https://github.com/vercel/next.js/blob/canary/contributing/repository/linting.md Co-authored-by: Delba de Oliveira <[email protected]>

Hi Team. This PR fixes an issue in the examples where the params object was destructured without being awaited. Co-authored-by: Delba de Oliveira <[email protected]>

#73269) ### Why? Follow-up of #73193, as we undeprecate revalidate APIs, the codemod should be reverted as well. Fortunately, there were no docs change. Reverts #72826

…mple (#73251) Improving Documentation - I noticed that using const instead of let in the example code for data fetching and caching is more appropriate, so I made the change. ![image](https://github.com/user-attachments/assets/73eb2dd7-386f-44ee-b820-7dbdd6a803eb) Co-authored-by: Delba de Oliveira <[email protected]>

…y in webpack loaders (#73112) ### What? * instead of depending on all env vars, track them via Proxy in webpack loaders * fix invalidation of evaluate pool after persistent cache restore

## Summary Follow up #45260. Update Kontent.ai brand name and demo link. ### Improving Documentation - [x] Run `pnpm prettier-fix` to fix formatting issues before opening the PR. - [x] Read the Docs Contribution Guide to ensure your contribution follows the docs guidelines: https://nextjs.org/docs/community/contribution-guide Co-authored-by: Delba de Oliveira <[email protected]>

### What This PR improves the reliability of `unstable_after` in a self-hosted context, i.e. on a long-running server (`next start`). when the server receives a signal to gracefully shutdown, we'll wait for any operations scheduled with `unstable_after` to finish, and then exit. ### Why It's less surprising than dropping pending operations on the floor if an instance is getting shut down. This mirrors the standard behavior of finishing handling pending requests before exiting -- we're just extending this a bit to include `after` tasks. ### How Previously `waitUntil` was just a noop in `next start`, so passing a promise to it did nothing. This PR changes the implementation to instead use an "awaiter" that stores promises passed to it and lets us wait for all of them to complete. We then use some existing `onCleanup` machinery to make sure that they're actually awaited before shutdown. Note that I've now made `NextServer.close()` execute before shutdown. This is notable because AFAICT this wasn't being called anywhere before, which is a bit suspicious. ### How (in a bit more detail) This is all a bit hard to follow so here's an overview of how it works: - We have two entrypoints: `startServer` and `NextCustomServer` - `startServer`: - creates a cleanup set (`AsyncCallbackSet`) - calls `createRequestHandlers`, which... - passes `onDevServerCleanup` to turbopack's hot reloader (previously `onCleanup`) - creates a `NextServer` - which will create a `NextNodeServer` - which sets up an `Awaiter` to ensure all `after` callbacks finished running. - adds a call to `NextServer.close()` to the cleanup set - when we receive SIGINT/SIGTERM, we run `NextServer.close()`, which runs `NextNodeServer.close()`, which waits for pending `after` callbacks to finish using the Awaiter. - `NextCustomServer`: - does basically the same setup with a cleanup set and `createRequestHandlers` - but it does not set up a SIGINT/SIGTERM handler to run the cleanups. instead, it expects the user to call its `cleanup()` method when appropriate (basically like we do in `startServer`) - i'm gonna do a follow up and update the custom server docs to reflect this - also going to try to warn if a cleanup was needed but the process exited without it calling it

Clean up a bad refactor

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[pull] canary from vercel:canary #188

[pull] canary from vercel:canary #188

Commits on Nov 13, 2024

Commits on Nov 14, 2024

Commits on Nov 15, 2024

Commits on Nov 16, 2024

Commits on Nov 17, 2024

Commits on Nov 18, 2024

Commits on Nov 19, 2024

Commits on Nov 20, 2024

Commits on Nov 21, 2024

Commits on Nov 22, 2024

Commits on Nov 23, 2024

Commits on Nov 24, 2024

Commits on Nov 25, 2024

Commits on Nov 26, 2024

Commits on Nov 27, 2024

[pull] canary from vercel:canary #188

Are you sure you want to change the base?

[pull] canary from vercel:canary #188

Commits on Nov 13, 2024

Commits on Nov 14, 2024

Commits on Nov 15, 2024

Commits on Nov 16, 2024

Commits on Nov 17, 2024

Commits on Nov 18, 2024

Commits on Nov 19, 2024

Commits on Nov 20, 2024

Commits on Nov 21, 2024

Commits on Nov 22, 2024

Commits on Nov 23, 2024

Commits on Nov 24, 2024

Commits on Nov 25, 2024

Commits on Nov 26, 2024

Commits on Nov 27, 2024