It should be possible to configure Cachet via a cachet:install Artisan command.

v2.x had an interactive command that would ask you questions and store the configuration.

For v3.x, this won't work exactly the same as we now have a mix of .env and database settings, but the idea is the same.

• Laravel version: 13.24.0
• PHP version: 8.5.9
• Database driver & version: SQLite 3 in memory, through Testbench 11.1

I was counting queries on an ordinary index endpoint and ended up with 101 queries for 100 records with ?include=comments. The same listing built as Post::with('comments')->get() gives 2. Before I read anything into that I'd like to know whether the loading is meant to be the developer's job here.

Where it comes from, as far as I can tell. ResolvesJsonApiElements:: compileResourceRelationships() calls $this->resource->loadMissing(...), and $this->resource in that method is always a single model, because both callers return early unless it is one. AnonymousResourceCollection::toAttributes() then maps resolveResourceData() over the collection item by item, so every item runs its own loadMissing against its own model.

The nested level behaves differently, and that's why I'm asking rather than just eager loading and moving on. With 10 posts and ?include=comments.author I get 21 queries: one for the posts, ten for the comments, ten for the authors. But the author lookups are batched inside each post, select * from "authors" where "authors"."id" in (1, 2, 3), because the second loadMissing runs on $relatedModels, which is already a collection. So the batching is there. It just never reaches the outer level.

What I measured, all on 13.24.0:

I'm leaving timings out on purpose. This is sqlite in memory, so the numbers there say more about serialisation than about the database, and the counts are what matters.

The case for "working as intended", which I can make myself: the call is loadMissing and not load, which reads like topping up whatever the developer forgot, not like owning the loading, and the documentation does show eager loading in the controller. If that's the answer then this is a documentation note at most and I'll drop it. It just seems like a lot to leave to a query log on an endpoint whose output is completely correct.

Two things I did not check, in case they matter: whether the same happens through an explicit ResourceCollection class and not the anonymous one, and whether chaperone() changes any of it. Both looked like separate paths when I skimmed them, so I might be missing something.

Steps to reproduce

Post hasMany Comment, Comment belongsTo Author. Resources generated with make:resource --json-api, PostResource declaring $relationships = ['comments'], CommentResource declaring $relationships = ['author'].

Route returning PostResource::collection(Post::all()), 100 posts with three comments each, then GET /posts?include=comments with Accept: application/vnd.api+json and DB::listen() counting.

symfony 7.4 php 8.5

Development is taking place on IIS, in the "Symfony" folder. Instead of the address "/Symfony/vendor/flasher/flasher.min.js" like other resources embedded via "asset," the address "/vendor/flasher/flasher.min.js" is inserted, which results in an error. The CSS file isn't embedded at all.

These messages aren't displayed: https://github.com/symfony/flex/blob/1.x/src/Configurator/DockerComposeConfigurator.php#L51

Octane Version

2.8.1

Laravel Version

10.48.28

PHP Version

8.3.16

What server type are you using?

Swoole

Server Version

6.0.1

Database Driver & Version

No response

Description

I'm using Envoyer to deploy a Laravel Octane application, taking advantage of its zero-downtime deployment features.

However, Octane currently does not support zero-downtime deployment because it does not follow symlink directories. It always remains in the actual directory (instead of the symlinked one) where the Octane server was initially started. When the outdated release directory is deleted, Octane continues running in that location, causing errors on every request and resulting in 500 response codes on the live application.

Error thrown:

Warning: require(/var/www/domain.com/releases/202502010200023/vendor/laravel/octane/bin/bootstrap.php): Failed to open stream: No such file or directory in /var/www/domain.com/releases/202502010200023/vendor/laravel/octane/bin/swoole-server on line 18
Fatal error: Uncaught Error: Failed opening required '/var/www/domain.com/releases/202502010200023/vendor/laravel/octane/bin/bootstrap.php' (include_path='.:/usr/bin/[email protected]/8.3.16/share/[email protected]/pear') in /var/www/domain.com/releases/202502010200023/vendor/laravel/octane/bin/swoole-server:18
Stack trace:
#0 /var/www/domain.com/releases/202502010200023/vendor/laravel/octane/bin/swoole-server(95): {closure}(Array)
#1 [internal function]: {closure}(Object(Swoole\Http\Server), 0)
#2 /var/www/domain.com/releases/202502010200023/vendor/laravel/octane/bin/swoole-server(170): Swoole\Server->start()
#3 {main}
  thrown in /var/www/domain.com/releases/202502010200023/vendor/laravel/octane/bin/swoole-server on line 18
#1 [internal function]: {closure}(Object(Swoole\Http\Server), 1)
#1 [internal function]: {closure}(Object(Swoole\Http\Server), 3)
#1 [internal function]: {closure}(Object(Swoole\Http\Server), 2)
#1 [internal function]: {closure}(Object(Swoole\Http\Server), 4)
#1 [internal function]: {closure}(Object(Swoole\Http\Server), 5)

Steps To Reproduce

Use any 0-downtime deployment or test it manually using the following instructions:

1. Use the Swoole driver as an example.
2. cd one directory up from the project's base path.
3. Create a current symlink directory for your project using the command: ln -nsf ./octane-project-test ./current
4. Start the Octane server: php ./current/artisan octane:start
5. Copy your project to another directory: cp -R ./octane-project-test ./octane-project-test-new
6. Activate the new release: ln -nsf ./octane-project-test-new ./current
7. Reload the Octane server: php ./current/artisan octane:reload
8. Remove the original project directory: rm -rf ./octane-project-test

This will probably be a component to search and select an image from unsplash.

Octane Version

2.13.1

Laravel Version

12.38.1

PHP Version

8.4.14

What server type are you using?

Swoole

Server Version

nginx

Database Driver & Version

No response

Description

When registering a singleton via AppServiceProvider, the constructor is triggered only once per worker, as expected. However, when using the attribute-based singleton registration (https://laravel.com/docs/12.x/container#singleton-attribute ), the constructor appears to be triggered once per request scope, behaving more like a scoped service. This may be due to the sandbox being registered earlier.

Steps To Reproduce

Check running this part

namespace App\TestService;

use Illuminate\Container\Attributes\Singleton;

#[Singleton]
class TestService
{
    public function __construct()
    {
        echo("__construct TestService");
        echo(spl_object_id($this));
    }
}

in some controler use

app(\App\TestService\TestService::class);
app(\App\TestService\TestService::class);
app(\App\TestService\TestService::class);

When you install a UX package, if you have AssetMapper installed, we importmap:require the packages you need. We should also importmap:remove those when the package is uninstalled.

https://github.com/symfony/recipes/issues/1089#issuecomment-1885459144

Description: I am using Flasher 2.0.1 in my Laravel 11 project and would like to know how to properly use the Flasher library in JavaScript. Additionally, I want to ensure that the JavaScript configuration is consistent with the settings defined in the flasher.php configuration file.

Details: Flasher Version: 2.0.1 Laravel Version: 11 Problem: Need guidance on using the Flasher library with JavaScript in a way that mirrors the configuration set in flasher.php.

I tried the example config from the docs to a Laravel/Inertia/Vue app, and it don't work.

If use the code (HandleInertiaRequests.php): 'messages' => flash()->render([], 'array'),

Error: Flasher\Prime\Flasher::render(): Argument #1 ($presenter) must be of type string, array given, called in D:\Laragon6\www\mobi-care\app\Http\Middleware\HandleInertiaRequests.php on line 36

And changing it to: 'messages' => flash()->render('array'), it loads the page, but, I only see the notifications if i press F5, with its not the desired state.

Octane Version

v2.3.10

Laravel Version

v11.6.0

PHP Version

v8.3.6

What server type are you using?

FrankenPHP

Server Version

v1.1.4 PHP 8.3.6 Caddy v2.7.6

Database Driver & Version

Postgress

Description

Last week we updated our app previously using php-fpm running on Forge to use Laravel Octane with FrankenPHP. Our site is mostly an API that handles analytics events (Like google analytics). It uses the default Laravel api throttling.

In staging our app worked fine (30 req/sec same IP), but when deploying to production (1400 req/sec, different IPs) it started to fail, giving a lot of 429 Too Many Requests.

I quickly rolled back to php-fpm and after a few hours tried again with the same problem. Rolled back and the next day I switched to Swoole and it worked perfectly without changing a single line of code nor having to redeploy anything. So I can confidently say that is NOT a bug in my code, but rather a bug with FrankenPHP or the Octane integration with FrankenPHP.

My theory is that the RateLimiter is not reseting between requests so it's shared between different users. So multiple different users trigger the rate limiter:

This is my Rate limiter configuration:

// AppServiceProvider

RateLimiter::for('api', function (Request $request) {
    return Limit::perMinute(60)->by($request->user()?->id ?: $request->ip());
});

our production CACHE_STORE is redis. Throttling worked perfectly fine without octane and with octane but using Swoole. It failed with hundred of 429 Too Many Requests after installing FrankenPHP.

This is our bootstrap/app.php:

<?php

use Illuminate\Foundation\Application;
use Illuminate\Foundation\Configuration\Exceptions;
use Illuminate\Foundation\Configuration\Middleware;
use Illuminate\Support\Facades\App;

return Application::configure(basePath: dirname(__DIR__))
    ->withRouting(
        commands: __DIR__.'/../routes/console.php',
        health: '/up',
        then: function () {
            Route::middleware('api')
                ->prefix('api')
                ->as('api.')
                ->domain(config('app.domain'))
                ->group(base_path('routes/api.php'));

            Route::middleware('web')
                ->domain(config('app.domain'))
                ->group(base_path('routes/web.php'));

            Route::middleware('web')
                ->domain(config('playsaurus.ads.domain'))
                ->group(base_path('routes/ads.php'));
        }
    )
    ->withMiddleware(function (Middleware $middleware) {
        $middleware->throttleApi();

        $middleware->redirectTo(
            guests: '/login',
            users: '/',
        );

        $middleware->web(append: [
            \App\Http\Middleware\HandleInertiaRequests::class,
            \Illuminate\Http\Middleware\AddLinkHeadersForPreloadedAssets::class,
        ]);

        $middleware->api(append: [
            \App\Http\Middleware\ConfigureLocale::class,
        ]);

        $middleware->alias([
            'localize' => \App\Http\Middleware\ConfigureLocale::class,
            'embed' => \App\Http\Middleware\AllowsEmbeding::class,
        ]);
    })
    ->withExceptions(function (Exceptions $exceptions) {
        $exceptions->dontReport([
            \App\Services\Announcements\InvalidVariantKey::class,
            \App\Exceptions\CouponRedeemException::class,
        ]);
    })->create();

Steps To Reproduce

It's difficult to reproduce. Because I can't test it in production because that would mean a lot of downtime for our users.

My theory is that it would be possible to reproduce from multiple different IPs. But since I don't have the means to test it, I don't know.

Octane Version

2.6.1

Laravel Version

11.41.3

PHP Version

8.3

What server type are you using?

Roadrunner

Server Version

2024.3.2

Database Driver & Version

No response

Description

ok this is in the borderline to be a bug but i write it anyway.

i am deploying using deployer , i suppose will happen for any Zero Downtime Deployments.

/releases/99/
/releases/100/
/releases/101/ 

and then a symbolic link from the latest release to /current , all pretty standard.

the problem come when running artisan octane:start vendor/laravel/octane/src/Commands/StartRoadRunnerCommand.php

will set roadrunner server with '-o', 'server.command='.(new PhpExecutableFinder)->find().','.base_path(config('octane.roadrunner.command', 'vendor/bin/roadrunner-worker')),

where base_path(...) will resolve to /releases/101/ not to /current

so in next deployment e.g. (/releases/102/) , when we run artisan octane:reload it will still use old /releases/101/ source code (i checked and yes if i go back to /releases/101/ and make a change it will be present after artisan octane:reload).

the solution is very simple, in config/octane.php you just need to add:

    'roadrunner'=>[
        'command' => env('OCTANE_ROADRUNNER_WORKER_PATH', base_path('vendor/bin/roadrunner-worker')),
    ]

OCTANE_ROADRUNNER_WORKER_PATH=../../current

would be nice is this added to config/octane.php , and maybe don't use base_path

or even better a option to provide the path in octane:start

Steps To Reproduce

deploy a octane laravel app using deployer , octane:reload will not work as expected

Sulu CMS separates its assets into assets/admin and assets/website, this means when installing Encore within a Sulu CMS install you will have to change a bunch of paths: https://docs.sulu.io/en/latest/cookbook/webpack-encore.html

This isn't my main issue, I think this is acceptable for an initial setup. But I do think we can improve what happens when you composer require a symfony bundle that provides stimulus controllers.

In the Sulu + Encore setup, the controllers.json file lives in assets/website/controllers.json. But since this path is hardcoded here:

https://github.com/symfony/flex/blob/441f671b86a1b651e969dbc380bfee05e004780f/src/PackageJsonSynchronizer.php#L166-L170

The controllers.json is not updated automatically and there's no output telling you to manually do this either. So you're left a little lost in what is still missing. Also, figuring out what to manually add in controllers.json is quite tricky since most existing UX bundles don't document this manual setup.

Before I start hacking away at a PR, any suggestions how we can solve this properly? Or do we accept that this is not configurable and those who use custom paths just have to deal with it?

Thank you in advance.

Immediately after installing php-flasher/flasher-laravel, the application breaks with the following error:

foreach() argument must be of type array|object, bool given Simply installing the package causes this fatal error.

Steps to Reproduce:

Create a fresh Laravel app.

Run:

composer require php-flasher/flasher-laravel

Load any page in the app — no usage of flasher() or flasher_render() yet.

Laravel crashes with a foreach() error.

Is LDAP support planned? I saw there was a lot of interest in this with version 2. Would be nice to have this onboard for v3. There is a package available https://github.com/DirectoryTree/LdapRecord-Laravel https://chrispian.com/laravel-filament-tutorial-customize-login-to-use-ldap

Hello, When using php-flasher-toastr with Laravel 11, I encounter a TypeError: array_replace_recursive(): Argument #2 must be of type array, int given error in Illuminate\Translation\FileLoader at line 130. My environment:

• PHP: 8.3.7
• Laravel: 11.4.2
• php-flasher/flasher-toastr-laravel: 2.1
• php-flasher/flasher-laravel: 2.1 The error disappears when I uninstall php-flasher/flasher-toastr-laravel and php-flasher/flasher-laravel. Alternatively, ensuring that the value passed to with() is always a string (e.g., with('success', 'Message')) prevents the error.

Hello,

Is it possible to close a flash message by clicking on it and not just on clicking on the cross? If so, how to achieve it?

Regards, Fred

using prompt engineering techniques provide enhancement to the Claude service class in the Claude prompt class, creating functions commonly used in Laravel applications. For example, brand builder, seo product optimizer, video and or image captioner, automated chatbots #goodfirstissue

Octane Version

2.9

Laravel Version

12

PHP Version

8.4

What server type are you using?

FrankenPHP

Server Version

1.12

Database Driver & Version

No response

Description

The current Caddy configuration template for FrankenPHP utilizes a broad try_files directive: try_files {path} frankenphp-worker.php

This logic causes all requests for non-existent static files (e.g., versioned assets like app-v123.js after a deployment) to be forwarded to the PHP worker. In high-traffic environments, a surge of requests for missing assets triggers unnecessary PHP process execution, leading to worker saturation, increased CPU/Memory overhead, and potential system failure.

Technical Impact

• Performance: High overhead for 404 responses that should be handled at the web-server level.
• Stability: Potential DoS condition when static asset paths are brute-forced or when client-side caches request deprecated versioned files.
• Resource Allocation: PHP workers are occupied by I/O-bound requests for missing files instead of processing dynamic application logic.

Proposed Solution Refactor the Caddyfile logic to isolate static file handling. A dedicated matcher for common file extensions should be implemented to ensure that if a static file is not found on disk, Caddy returns a 404 Not Found immediately, bypassing the php_server or php_dispatch logic.

Recommended Configuration Change Define a matcher for static assets and handle them via file_server exclusively:

@static {
    path *.css *.js *.mjs *.map *.ico *.png *.jpg *.jpeg *.gif *.svg *.webp *.avif *.woff *.woff2 *.ttf *.otf *.eot *.mp4 *.webm *.ogg *.mp3 *.wav *.pdf *.zip *.txt *.xml *.json *.wasm *.webmanifest *.manifest
}

file_server @static

Steps To Reproduce

1. Deploy Laravel Octane with FrankenPHP.
2. Request a non-existent file with a .js or .css extension.
3. Observe that the request is processed by frankenphp-worker.php (returning a Laravel 404) instead of a native Caddy 404.

The dashboard should offer the ability to toggle the theme between:

• Automatic (System)
• Light
• Dark

Further, the dashboard should have a "Theme" option to force a theme mode.

On the record update for an active incident, there needs to be a tie to the component to be update the current component status.

Provide updates to new functionalities in the v2 of the package.

What about allowing displaying post-install messages again, eg when running this?

composer recipes the/package

Hi Developer,

I appreciate your work on this project—it's really well-built and helpful for developers like me. I recently implemented it and found it working great.

However, I noticed that the .fl-wrappe class currently has z-index: 10, which causes it to appear under the header when using a fixed or sticky header. I suggest updating it to z-index: 99999 to ensure proper visibility.

Thanks for your efforts and for sharing this project!

Best regards, Md. Jahangir Alam Rohan.

Octane Version

2.11.0

Laravel Version

12.20.0

PHP Version

8.3.23

What server type are you using?

FrankenPHP

Server Version

1.8.0

Database Driver & Version

No response

Description

I have an octane app running inside a docker container and use a bind mount for the application code. When editing files, fs events are not sent across the bind mount to the container, so the watcher does not detect them. I have the same issue reported in #487. There is also an open issue on the frankenphp repository https://github.com/php/frankenphp/issues/1616.

A potential solution would be to add an option to octane:frankenphp, like --chokidar to use chokidar instead of frankenphp's watcher.

Steps To Reproduce

Same as #487, but using frankenphp.

I would love to see support for oAuth in Catchet. I know there are some other (closed) issues requesting the same functionality, but I think this would be a great addition for Catchet.

NotOrm Package

The goal is to make abstract code based on the package notorm to build a new orm system

Hi 👋🏼!

I am coming here to gather some feedback on my idea before starting working on it.

Background

I wanted to create a POC of https://github.com/symfony/skeleton made for Sylius. I created a simple recipe for sylius/core-bundle, then an example skeleton repo and I have found out my recipes does not work as another recipe already write files with the same name. Then, I noticed symfony/framework-bundle is always put as the first recipe to be executed, and this is a thing I wish to be able to configure.

Goal

Somehow allow myself to make (in this POC case) sylius/core-bundle as a first recipe to be executed. Of course, I can fork symfony/flex, but it would be perfect to avoid this way.

Idea

The idea is simple, we allow configuring such list for example in this way:

{
    ...
    "extra": {
        "flex": {
            "prioritized-recipes": [
                "sylius/core-bundle",
                "another/sylius-package",
                ...
            ]
        }
    }
    ...
}

In Flex we could implement this +/- this way:

        // symfony/framework-bundle recipe should always be applied first after the metapackages
        // however, we allow to override it with a list of prioritized recipes
        $recipes = $this->getPrioritizedRecipes();
        $recipes = array_merge($recipes, [
            'symfony/framework-bundle' => null,
        ]);
        $packRecipes = [];
        $metaRecipes = [];

instead current

        // symfony/framework-bundle recipe should always be applied first after the metapackages
        $recipes = [
            'symfony/framework-bundle' => null,
        ];
        $packRecipes = [];
        $metaRecipes = [];

Why?

• In some projects, we may want to load our recipes before the Symfony's ones
• In frameworks based on Symfony (like Sylius) we need to set up the whole project in our own way, so framework-bundle as a first recipe to be executed makes it unable for us

Other options

I have not checked it yet, but I believe we can achieve the similar feature using Composer's Event Dispatcher. But first, I would like to hear if such a feature is welcomed. Or maybe you have a better idea how to solve this. I am open to provide such feature right after we agree on some solution.

using prompt engineering techniques provide enhancement to the Gemini service class in the Gemini prompt class, creating functions commonly used in Laravel applications.

We need a method of migrating users from 2.4 to 3.x.

Things to consider:

• Migrate settings
• Migrate resources
• Migrate subscriptions / subscribers
• Migrate API Keys

AOB (please add to this list)

As a comment on the latter, once #183 is merged, we'll need to programmatically migrate users 2.4 API Keys to 3.x using something like the following

$v2ApiKey = $v2User->api_key;

$v3User->tokens()->create([
    'name' => 'v2.4 API Key',
    'token' => hash('sha256', $v2ApiKey),
    'abilities' => ['*'],
    'expires_at' => now()->addDays(90),
]);

@jbrooksuk We need to decide on the abilities and expires_at. IMO it makes sense to throw something into the upgrade documentation that old 2.4 API Keys will continue to work for up to 90 days. But if they want to use the API after that, then they'll need to generate a new API Token using the 3.x dashboard.