All Issues
24,993 verified issues
failed to get console mode for stdout: The handle is invalid.
Description of the issue The `docker-compose build` command prints this `failed` error every time. Context information (for bug reports) Output of `docker-compose version` [code block] Output of `docker version` [code block] Output of `docker-compose config` (Make sure to add the relevant `-f` and other flags) [code block] and the Dockerfile can be anything, this works [code block] Steps to reproduce the issue 1. Create any Dockerfile on Windows 2. Create any Docker Compose manifest with a build section pointing to the same Dockerfile 3. Run `docker-compose build` Observed result The Docker Compose build command always prints the `failed` line. Nothing else _seems_ to be wrong, but I'm not sure. [code block] Expected result The plain Docker build command works as expected. [code block] Stacktrace / full error message [code block] Additional information OS version / distribution, `docker-compose` install method, etc.
"http: Hijack is incompatible with use of CloseNotifier in same ServeHTTP call" errors
Using 1.8-rc1 and docker 1.12-rc3 with the following compose file: [code block] and Dockerfile: [code block] If I do: [code block] Looks like a missed case in https://github.com/docker/compose/pull/1374 cc @dnephin
(FR): Implement 'minimumReleaseAge' as a security feature (like pnpm added recently)
Ref: https://github.com/pnpm/pnpm/issues/9921 https://pnpm.io/settings#minimumreleaseage https://socket.dev/blog/pnpm-10-16-adds-new-setting-for-delayed-dependency-updates https://github.com/raineorshine/npm-check-updates/issues/1532 Copying an excerpt from one of these articles: > Following a wave of high-profile supply chain attacks targeting popular npm packages, pnpm has shipped a new minimumReleaseAge setting in version 10.16 that delays installation of newly published packages.
JS exceptions thrown while just idling (WebSocket is already in CLOSING or CLOSED state.)
You want to: [x] report a bug [ ] request a feature Current behaviour While leaving a socket connected, but doing no other activity in my program, I see intermittently (but fairly consistently) an exception thrown in browser console, from inside of the socket.io machinery (specifically in `backo2/index.js` line 83. That error is: [code block] Steps to reproduce (if the current behaviour is a bug) I have two tabs open with client sockets connected to the same server (localhost) via https. Both clients are sitting idle with nothing else happening in browser or server (except for whatever keep-alive pings socket.io is doing). Both of them are joined to a single channel (via `join(..)` on the server). Otherwise, nothing else special. Here's how I create the server socket instance: [code block] And in the client: [code block] Expected behaviour I expect disconnections and reconnections from time to time, but I don't expect spurious JS exceptions thrown by the library when I'm not doing anything else over the connections. Setup - OS: Mac OSX - browser: Chrome 66, node 10.2.1 - socket.io version: 2.1.1 Other information (e.g. stacktraces, related issues, suggestions how to fix) Expanded stack trace: [code block]
[DevTools Bug]: Blank tools localhost only
Website or app google.com Repro steps This started after last update 4.20.0 How often does this bug happen? Every time DevTools package (automated) _No response_ DevTools version (automated) _No response_ Error message (automated) _No response_ Error call stack (automated) _No response_ Error component stack (automated) _No response_ GitHub query string (automated) _No response_
docs embedded in object w/ arbitrary keys?
Is there a way to use embedded docs within a object that doesn't have predetermined keys? i.e. Rather than having a situation like this: [code block] Being able to say something like: [code block] Makes it a little more array like in that way - but the idea is that I don't know the keys in advance, but I'd still like to be able to have an embedded doc.
instance.defaults changes global object
If I have two instances of axios created with axios.create and I set a default header on one, it also changes it on all the other ones in existence. [code block] Both instance1 and 2 now have the Authorization header changed. I believe the problem is that the util.merge operation doesn't do deep merging, though as you see below, only if the result[key] and val are both objects, does it do a recursive merge, otherwise it assigns the value by reference. [code block]
[Introspection] Cross Schema Foreign Keys
`relational.fit.cvut.cz` DB server has 2 schemas `ctu_crossSchema1` and `ctu_crossSchema2` that reference each other in foreign keys: Error messages during introspection: [code block] [code block] SQL: [code block]
[NEXT-378] Next 13: Navigation with Link does not scroll up the page
Verify canary release [X] I verified that the issue exists in the latest Next.js canary release Provide environment information Operating System: Platform: linux Arch: x64 Version: #1 SMP Mon Sep 19 19:14:52 UTC 2022 Binaries: Node: 16.15.0 npm: 8.5.5 Yarn: 3.2.4 pnpm: N/A Relevant packages: next: 13.0.2 eslint-config-next: 13.0.0 react: 18.2.0 react-dom: 18.2.0 What browser are you using? (if relevant) Chrome 107.0.5304.87 How are you deploying your application? (if relevant) Vercel Describe the Bug When navigating between pages using `Link`, the page does not scroll up by default wich is supposed to be the default behaviour of `next/link` Expected Behavior The page scrolls up when switching pages unless I manually disable this feature of `next/link` Link to reproduction https://beta.otpfy.com/ To Reproduce Access https://beta.otpfy.com/ Scroll down until you see the "Get started for free" button Click the given button * Check that when navigating to the `auth` page the scroll is preserved and not scrolled to the top \: <sub>From SyncLinear.com | NEXT-378</sub>
Break on multiple chained calls
I've mentioned this possibility in a few different issue (#1099, #1282, #1401), but I thought a separate issue might be appropriate. 😄 I'd like for prettier to break on multiple chained calls. I prefer a higher `printWidth` than 80 in most cases (such as strings, log statements, class declarations etc), but increasing it leads prettier to one-line a lot of chained expressions that were multi-lined. Eslint has a rule for this called `newline-per-chained-call`. It defaults to 3 chained calls should break, which sounds like a sane starting point. This rule is unfixable, unfortunately, but Prettier could fix it! 😀 Even for the default width of 80 some things are inlined which IMO are more readable expanded. I don't think there's any doubt which of these two are the easiest to read (especially considering the git diff scenario from eslint's docs): https://prettier.github.io/prettier/#%7B%22content%22%3A%22d3%5Cn%20%20.select(%5C%22body%5C%22)%5Cn%20%20.selectAll(%5C%22p%5C%22)%5Cn%20%20.data(%5B1%2C%202%5D)%5Cn%20%20.enter()%5Cn%20%20.style(%5C%22color%5C%22%2C%20%5C%22white%5C%22)%3B%5Cn%5Cnd3%5Cn%20%20.select(%5C%22body%5C%22)%5Cn%20%20.selectAll(%5C%22p%5C%22)%5Cn%20%20.data(%5B1%2C%202%2C%203%5D)%5Cn%20%20.enter()%5Cn%20%20.style(%5C%22color%5C%22%2C%20%5C%22white%5C%22)%3B%5Cn%22%2C%22options%22%3A%7B%22printWidth%22%3A80%2C%22tabWidth%22%3A2%2C%22singleQuote%22%3Afalse%2C%22trailingComma%22%3A%22none%22%2C%22bracketSpacing%22%3Atrue%2C%22jsxBracketSameLine%22%3Afalse%2
Image Issue on Nginx
HI, Image works in development and production (only locally, not remotely), but in nginx I get this error (from nginx console): [code block] my folders public/assets/icons/gdpr.png [code block] i try to convert the image in webp.. same error restart server.. same and this: [code block] i receive from chrome [code block] <img width="265" alt="image" src="https://github.com/vercel/next.js/assets/54176220/5dfa8d65-6d90-4c65-8355-7da4ed560334"> Any ideea? Thanks! Additional information [code block] <sub>NEXT-1680</sub>
feat: Support react-query@5 and edge
> ℹ️ Note from maintainers > > We have support for tanstack query 5 in our new major - see @juliusmarminge comment here > > You can upgrade using something akin to > > [code block] Describe the feature you'd like to request So `react-query` just had a new major version bump which removes some react kludge, which hopefully will make it tRPC + query compatible with edge runtimes (i.e. CF workers, Vercel edge etc.) Describe the solution you'd like to see https://tanstack.com/query/v5/docs/react/guides/migrating-to-v5 describes migration steps. Would need some updates to `@trpc/next`. Describe alternate solutions I believe there's nothing inherently breaking on edge runtimes in `@trpc/next`, so it's up to `react-query`. Additional information _No response_ 👨👧👦 Contributing - [ ] 🙋♂️ Yes, I'd be down to file a PR implementing this feature! <sub>TRP-25</sub>
Supabase down on JIO ISP
Bug report - [x] I confirm this is a bug with Supabase, not with my own application. - [x] I confirm I have searched the Docs, GitHub Discussions, and Discord. Describe the bug All Supabase API requests (`.supabase.co`) are completely inaccessible for users on Jio (Reliance) — The ISP appears to be performing DNS poisoning on `.supabase.co` domains, resolving them to a Jio-owned sinkhole IP (`49.44.79.236`) instead of the actual AWS IPs. This results in `ERR_CONNECTION_TIMED_OUT` for all API calls, Edge Functions, and Auth endpoints. This is not a project-specific issue — all Supabase projects are affected. The Supabase marketing site (`supabase.com`) loads fine (different domain), but `supabase.co` (used for all API/project endpoints) is fully blocked. This affects a massive portion of Indian internet users and could silently break any production app using Supabase for users on Jio broadband or Jio mobile data. To Reproduce 1. Connect to any Jio network in India (broadband or mobile data) 2. Open any app that makes API calls to `.supabase.co` (e.g., a project using Supabase Auth, REST API, or Edge Functions) 3. API requests fail with `ERR_CONNECTION_TIMED_OUT` 4. The same app works perfectly on other ISPs (Airtel, Vi, BSNL) or through a VPN DNS resolution proof: [code block] The resolved IP `49.44.79.236` belongs to Jio/Reliance, not to AWS (where Supabase is hosted). This confirms DNS-level blocking/poisoning. curl output: [code block] Both `supabase.co` and p
Can't post FormData since Axios 0.25.0
Describe the issue I can't post FormData since `"axios": "^0.25.0",` `Error: Request failed with status code 400` Example Code [code block] Expected behavior, if applicable That it works Environment - Axios Version: 0.25.0 - Node.js Version: v14.17.6 - OS: iOS 15, Android 11 (But on all platforms and versions) - Additional Library Versions: React Native 0.64.3 Additional context/Screenshots No problem with `"axios": "^0.24.0",` I saw this PR https://github.com/axios/axios/pull/3757 on 0.25.0, but I don't know what I have to change in code. Thanks for your help 👍
App crashes randomly on Android 12 with `JSApplicationIllegalArgumentException: Animated node with tag ### does not exists`
Description com.facebook.react.bridge.JSApplicationIllegalArgumentException: Animated node with tag 201 does not exists at com.facebook.react.animated.NativeAnimatedNodesManager.connectAnimatedNodes(NativeAnimatedNodesManager.java:317) at com.facebook.react.animated.NativeAnimatedModule$15.execute(NativeAnimatedModule.java:681) at com.facebook.react.animated.NativeAnimatedModule.executeAllOperations(NativeAnimatedModule.java:258) at com.facebook.react.animated.NativeAnimatedModule.access$400(NativeAnimatedModule.java:85) at com.facebook.react.animated.NativeAnimatedModule$3.execute(NativeAnimatedModule.java:287) at com.facebook.react.uimanager.UIViewOperationQueue$UIBlockOperation.execute(UIViewOperationQueue.java:573) at com.facebook.react.uimanager.UIViewOperationQueue$1.run(UIViewOperationQueue.java:908) at com.facebook.react.uimanager.UIViewOperationQueue.flushPendingBatches(UIViewOperationQueue.java:1019) at com.facebook.react.uimanager.UIViewOperationQueue.access$2600(UIViewOperationQueue.java:47) at com.facebook.react.uimanager.UIViewOperationQueue$DispatchUIFrameCallback.doFrameGuarded(UIViewOperationQueue.java:1079) at com.facebook.react.uimanager.GuardedFrameCallback.doFrame(GuardedFrameCallback.java:29) at com.facebook.react.modules.core.ReactChoreographer$ReactChoreographerDispatcher.doFrame(ReactChoreographer.java:175) at com.facebook.react.modules.core.ChoreographerCompat$FrameCallback$1.doFrame(ChoreographerCompat.java:85) at androi
[BUG]: Many TypeScript Compile Errors (drizzle-orm/neon-http ; Postgres)
What version of `drizzle-orm` are you using? ^0.28.6 What version of `drizzle-kit` are you using? ^0.19.13 Describe the Bug During the build step for Typescript, I run into the same "10 errors in 9 files" within the drizzle-orm Node modules package. Reproduced in 3 different projects, one of new was fresh, following the instructions step-by-step for Neon integration over http, as Firebase Cloud Functions do not support Web Sockets (to my knowledge; I could be wrong.) Step 1: [code block] Step 2: Ensure your env contains the string you need and is being read, then - [code block] Step 3: [code block] Build Errors (Same In all Projects Tested): [code block] Expected behavior Expect TypeScript to be able to build. Confirmed errors exist in individual node_module files noted in build process as well Environment & setup Server Platform: Firebase Cloud Functions Node Version: 18 Database Type: Neon.tech PostgreSQL Connection Type: http, ssl TypeScript Version: "^4.9.0" Editor: VSCode
Routes using data fetching methods return 304 on non-conditional requests
Link to the code that reproduces this issue https://github.com/kevineinarsson/next-revalidate-issue To Reproduce 1. Start the application in production mode (build & start) 2. Curl one of the routes: [code block] These should return 200:s, no matter how many times you curl. Note the x-nextjs-cache header which indicates whether a revalidation of the content is performed (due to fetch options including `next: { revalidate: 5 }`) [code block] This should return a 304, since we're matching against an ETag and the content matches. However, if a revalidation occurs from a request that includes a If-None-Match header, the `.meta` will now contain `status: 304`, causing subsequent requests without an ETag to return 304 not modified, which should not happen. Current vs. Expected behavior Cache updates should not store conditional headers such as 304. The result of the operations above should be a 200 in the metadata file. Verify canary release - [X] I verified that the issue exists in the latest Next.js canary release Provide environment information [code block] Which area(s) are affected? (Select all that apply) Not sure Additional context Issue does not appear on 13.4.17. In dynamic routes, removing `generateStaticParams()` makes the problem disappear, but doing so leads to a response without an ETag. Using `export const dynamic = 'force-dynamic';` works for static pages, but of course makes them dynamic. <sub>NEXT-1699</sub>
Custom class name completion contexts
There has been quite a few requests for the extension to support class name completions in contexts other than a standard class(Name) attribute. Some examples: - `tailwind.macro` and `twin.macro` (#46, tailwindlabs/tailwindcss-intellisense#123): [code block] - `tailwind-rn` (#99): [code block] - `tailwind-ppx` (#120): [code block] I just wanted to consolidate all of these requests into a single issue that can be tracked more easily, as I think the solution could be the same for each of them. I am reluctant to hard-code each of these cases into the extension because none of them are "official" methods of using Tailwind, and it may become a maintenance burden. However, I am open to the idea of adding a user setting which would allow the definition of custom regular expressions. For example for `tailwind-rn` your regular expression might be something like: `/\btailwind\([^)]+/ig` If you're interested in this feature feel free to "watch" this issue for updates, and post any comments/suggestions you may have.
Wrong indentation on ternary operator
I can't reproduce it on http://mib.im/try-prettier/, but if I run: [code block] On my Popper.js, the file `src/tooltip/index.js` at line 59 will have: [code block] Note that the `return` line has wrong indentation.
Problem parsing body: No signatures found matching the expected signature for payload
I'm using AWS API Gateway/Lambda which passes the request body as an actual JSON object to my Lambda function, where I am calling `stripe.webhooks.constructEvent` on it. It seems this function expects the payload to be stringified JSON, so I run `JSON.stringify` on the request body, but this fails to sign (I'm guessing because JSON.stringify subtly changes what the JSON string would look like coming from the server). Is it possible for constructEvent to detect if the payload is a JSON object and skip parsing it? I mean, in a way that maintains security?