Found the \Illuminate\Database\Eloquent\Concerns\HasAttributes::originalIsEquivalent() method during code diving and it should be usable for our trait! 🎉 This PR would be mainly reasearch when that method was added, if it's compatible with our version support rules and switching our own change detection logic with the core one.

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.

Hello,

I noticed that there is an incompatibility with the mbstring polyfill and PHP 8.1 / Alpine Linux, which breaks a lot of my projects as soon as the php81-mbstring is not installed, but php81-iconv is installed:

Example:

Warning: iconv(): Wrong encoding, conversion from "ASCII" to "UTF-8//IGNORE" is not allowed in phar:///var/www/localhost/htdocs/phpstan.phar/vendor/symfony/polyfill-mbstring/Mbstring.php on line 736

It looks like //IGNORE is not accepted since echo iconv('UTF-8', 'UTF-8', 'test'); works, while echo iconv('UTF-8', 'UTF-8//IGNORE', 'test'); doesn't

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.

Describe the bug I have a table which has an order column. When I reorder the elements I update each model with the new order. Because the table has some large JSON columns I select only id and order columns for efficiency. I set the new order and save the model.

foreach (Faq::query()->get(['id', 'order']) as $faq) {
    $faq->order = $newOrder[$faq->id];
    $faq->save();
}

The diffs created in the activity_log table look like this:

{
   "old": {
      "order": 1,
      "answer": null,
      "question": null
   },
   "attributes": {
      "order": 2,
      "answer": "long JSON content",
      "question": "long JSON content"
   }
}

From what I see laravel-activitylog fetches the model with all columns from the database before creating a log entry which kind of defeats the purpose of my "select list optimization" and produces an incorrect diff.

Generally I want these large JSON columns to be logged in case a user changes them but in this case I only change the order column so I would like to see only the order column in the diff.

I have tried running these updates straight off the Eloquent builder:

Faq::query()->where('id', $id)->update(['order' => $newOrder[$id]]);

but in this case nothing is logged.

I have the following activitylog configuration for my models:

public function getActivitylogOptions(): LogOptions
{
    return LogOptions::defaults()
                     ->logAll()
                     ->logOnlyDirty()
                     ->logExcept([
                         'id',
                         'created_at',
                         'updated_at',
                     ]);
}

To Reproduce Select a subset of columns from the database, change these columns, save the model and see that other (not initially selected) columns are present in the diff.

Expected behavior Only columns that've actually been changed should be present in the diff.

Versions

• PHP: 8.1
• Database: MySQL 8
• Laravel: 9.31.0
• Package: 4.6.0

What happened?

After installation, I noticed that in the browser console it outputs an error of wireEl is undefined.

How to reproduce the bug

Just install this package in a livewire project, preferable with Filament's admin package. And see the browser console.

composer create-project laravel/laravel test;
composer require filamentphp/filament;
composer require spatie/laravel-blade-comments;

Then go through the browser console and see the error

Package Version

1.0.1

PHP Version

8.2.0

Laravel Version

10.13.0

Which operating systems does with happen with?

macOS

Notes

Filament ......................................................
Packages ...... filament, forms, notifications, support, tables
Version .............................................. v2.17.44
Views ..................................... PUBLISHED: filament

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

Hi! If I use this package with my project I has a problem. My package.json has a "git dependency" like this:

{
  "peerDependencies": {
    "package": "git+ssh://[email protected]/team/project.git"
  }
}

And npm-install-peers breaks.

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.

Not PHPUnit 11 because we still need to support for PHP 8.1

e.g. at the moment we have stuff like:

1. Symfony\Flex\Tests\Configurator\CopyFromPackageConfiguratorTest::testConfigureAndOverwriteFiles The at() matcher has been deprecated. It will be removed in PHPUnit 10. Please refactor your test to not rely on the order in which methods are invoked.

Help wanted 🙏

Currently you can only see the latest 3 articles, latest 5 threads and latest 5 replies of a user on their profile. It would be cool if we could paginate through those results. We already use livewire so would be cool if we could add some magic with that.

Laravel Version

12.*

PHP Version

8.*

Database Driver & Version

No matter

Description

As we all know, whereBetween method may accept CarbonPeriod as $values argument. Lets look inside:

    public function whereBetween($column, iterable $values, $boolean = 'and', $not = false)
    {
        $type = 'between';

        if ($values instanceof CarbonPeriod) {
            $values = [$values->getStartDate(), $values->getEndDate()];
        }

        $this->wheres[] = compact('type', 'column', 'values', 'boolean', 'not');

        $this->addBinding(array_slice($this->cleanBindings(Arr::flatten($values)), 0, 2), 'where');

        return $this;
    }

But CarbonPeriod may not have end date! We can construct DatePeriod two ways: passing end-date or passing number of intervals:

$period1 = new \DatePeriod(now(), new \DateInterval('P2M'), now()->addMonths(2));
$period2 = new \DatePeriod(now(), new \DateInterval('P2M'), 2);

$period1 has end-date, but $period2 hasn't.

So I think that in such case method should calculate end-date itself.

Next problem is how between works. Mysql documentation says that between is equivalent to the expression (min <= expr AND expr <= max). Now look at DatePeriod documentation. The fourth argument is an $options, it may accept DatePeriod::EXCLUDE_START_DATE and DatePeriod::INCLUDE_END_DATE.

If DatePeriod excludes start-date, the query should be (min < expr AND expr <= max). If DatePeriod excludes end-date, the query should be (min <= expr AND expr < max). If DatePeriod excludes both dates, the query should be (min < expr AND expr < max).

So at the moment Eloquent is not absolutely correct.

And the last but not the least: why this method accepts CarbonPeriod and not accepts base DatePeriod?

I think, that whereBetween

• should accept DatePeriod
• should evaluate end-date when required
• should respect EXCLUDE_START_DATE and INCLUDE_END_DATE options

Steps To Reproduce

No steps required

If a pack lists things in its "replace" section, unpacking this pack should copy these rules into the "replace" section of the root composer.json.

I think that implementing this requires patching the Unpacker class only.

Anyone up to give it a try?

There's no Expectation to expect a class to use HasRequestPagination rather than HasPagination.

Bug description

Currently, the plugin calculates width/height incorrectly when swapping aspect ratios, leading to incorrect HTML output. When changing the aspect ratio of an image, the longest side should remain fixed, and the other side should be calculated accordingly.

How to reproduce

For an image of 3000x1688px as a source:

{{ responsive:image ratio="16/9" md:ratio="9/16" }}

it returns:

<picture>
    <source [...] >
    <source [...] >
    <source [...] >
    <img src="/img/asset/hash/image.jpg?w=3000&amp;h=1688&amp;s=9358bb54be5d388a9a2529a4196dcfc6" alt="" width="3000" height="1688">
</picture>

Now for the same image as source but with:

{{ responsive:image ratio="9/16" md:ratio="16/9" }}

it returns:

<picture>
    <source [...] >
    <source [...] >
    <source [...] >
    <img src="/img/asset/hash/image.jpg?w=3000&amp;h=5333&amp;s=6c5c0bb8447dbd7ff61a692f98b5f46d" alt="" width="3000" height="5333">
</picture>

The calculated aspect ratio is correct, but the dimensions are not. The generated values exceed the original image size.

In this last example, the correct height should be 3000px, and the width should be 1688px to ensure the image stays within its original dimensions.

Environment

Environment
Application Name: local
Laravel Version: 11.41.3
PHP Version: 8.3.14
Composer Version: 2.8.5
Environment: development
Debug Mode: ENABLED
URL: local.test
Maintenance Mode: OFF
Timezone: Europe/Paris
Locale: en

Cache
Config: NOT CACHED
Events: NOT CACHED
Routes: NOT CACHED
Views: CACHED

Drivers
Broadcasting: null
Cache: file
Database: sqlite
Logs: stack / single
Mail: smtp
Queue: redis
Session: file

Statamic
Addons: 3
Sites: 1
Stache Watcher: Disabled (auto)
Static Caching: Disabled
Version: 5.46.0 PRO

Statamic Addons
rias/statamic-redirect: 3.9.3
spatie/statamic-responsive-images: 5.2.1
statamic/seo-pro: 6.5.0

Installation

Fresh statamic/statamic site via CLI

Antlers Parser

regex (default)

When attempting to filter tests by test or file name pressing the "Enter" key has no effect.

Pressing "Enter" at the default screen correctly triggers a "test run" where all tests are run.

• standard tags
• input fields
• components
• patterns

Add code snippets for each block

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

I would be good to support normalizer_get_raw_decomposition function that appeared in PHP 7.3. I've tried to implement it but seems that existing decomposition data is already optimized to get the final decomposition. Example test case is available here. Pinging @nicolas-grekas as he is the author of Normalizer polyfill.

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

Add in PHP 7.3 as Normalizer::normalize() argument for NFKC_Casefold normalization.

When encountering nested tokens, they are parsed only during formatting, not when compiling the pattern. And the type of tokens are validated only during the rendering as well. But the native implementation detects such error during the instantiation: https://3v4l.org/Sf69D This means that the error handling required for the polyfill is not the same than for the native implementation.

Hi,

With the polyfill, var_dump(mb_strlen(chr(254))) return 0. With the php8.0-mbstring extension, var_dump(mb_strlen(chr(254))) return 1;

versions : * v1.26.0

Thanks, Alex

Hello,

if you plan an scheduled maintenance with an START and END date it will never appear on the dashboard. Only if you create them with an START date only it will appear. I think this is a bit confusing, because if i plan an schedule an weekly update, i want to declare an timeslot and dont want to enter again the cachet-scheduled-maintenance to enter an date at the end of my maintenance.

Can this function be controlled or changed?

Thanks

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.

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

When calling mb_convert_encoding() with $fromEncoding === 'HTML-ENTITIES', the polyfill does not return functionally equivalent strings to the native function. This is because mb_convert_encoding() uses html_entity_decode() when $fromEncoding === 'HTML-ENTITIES' and that function does not return characters for many numeric entities 0-31 and 127-159. For example:

<?php

require "vendor/symfony/polyfill-mbstring/Mbstring.php";

use Symfony\Polyfill\Mbstring as p;

for($i = 0; $i < 1024; $i++) {
	$string = "&#" . $i . ";";
	$mbstring = mb_convert_encoding($string, 'UTF-8', 'HTML-ENTITIES');
	$polyfill = p\Mbstring::mb_convert_encoding($string, 'UTF-8', 'HTML-ENTITIES');
	if($mbstring != $polyfill) {
		echo "Mismatch: $string - mbstring: $mbstring; polyfill: $polyfill\n";
	}
}

outputs:

Mismatch: � - mbstring: ; polyfill: �
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring:; polyfill: 
Mismatch:  - mbstring:
                            ; polyfill: 
Mismatch:  - mbstring:
                            ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring:  polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch: ' - mbstring: '; polyfill: '
Mismatch:  - mbstring: ; polyfill: 
Mismatch: € - mbstring: €; polyfill: €
Mismatch:  - mbstring: ; polyfill: 
Mismatch: ‚ - mbstring: ‚; polyfill: ‚
Mismatch: ƒ - mbstring: ƒ; polyfill: ƒ
Mismatch: „ - mbstring: „; polyfill: „
Mismatch: … - mbstring: …; polyfill: …
Mismatch: † - mbstring: †; polyfill: †
Mismatch: ‡ - mbstring: ‡; polyfill: ‡
Mismatch: ˆ - mbstring: ˆ; polyfill: ˆ
Mismatch: ‰ - mbstring: ‰; polyfill: ‰
Mismatch: Š - mbstring: Š; polyfill: Š
Mismatch: ‹ - mbstring: ‹; polyfill: ‹
Mismatch: Œ - mbstring: Œ; polyfill: Œ
Mismatch:  - mbstring: ; polyfill: 
Mismatch: Ž - mbstring: Ž; polyfill: Ž
Mismatch:  - mbstring: ; polyfill: 
Mismatch:  - mbstring: ; polyfill: 
Mismatch: ‘ - mbstring: ‘; polyfill: ‘
Mismatch: ’ - mbstring: ’; polyfill: ’
Mismatch: “ - mbstring: “; polyfill: “
Mismatch: ” - mbstring: ”; polyfill: ”
Mismatch: • - mbstring: •; polyfill: •
Mismatch: – - mbstring: –; polyfill: –
Mismatch: — - mbstring: —; polyfill: —
Mismatch: ˜ - mbstring: ˜; polyfill: ˜
Mismatch: ™ - mbstring: ™; polyfill: ™
Mismatch: š - mbstring: š; polyfill: š
Mismatch: › - mbstring: ›; polyfill: ›
Mismatch: œ - mbstring: œ; polyfill: œ
Mismatch:  - mbstring: ; polyfill: 
Mismatch: ž - mbstring: ž; polyfill: ž
Mismatch: Ÿ - mbstring: Ÿ; polyfill: Ÿ

While many of these are control characters (and the native function does return them), the single quote (dec 39) is particularly problematic.

Laravel Version

12.41.1

PHP Version

8.5.0

Database Driver & Version

PostgreSQL 18.1 (pdo_pgsql, libpq 18.1) on ARM64 Docker

Description

Laravel crashes with SIGSEGV (exit code 139) when using SESSION_DRIVER=database with PostgreSQL on PHP 8.5. The crash occurs during session regeneration after Auth::attempt() or Auth::login().

Important: A standalone custom PDO session handler does NOT crash. The issue is specific to Illuminate\Session\DatabaseSessionHandler.

Server Logs

laravel.test-1 | WARN exited: php (exit status 139; not expected) laravel.test-1 | INFO reaped unknown pid 228 (terminated by SIGSEGV)

What Works

• SESSION_DRIVER=redis with PostgreSQL
• SESSION_DRIVER=file with PostgreSQL
• Custom PDO session handler (standalone)
• All regular PDO PostgreSQL queries

What Doesn't Work

• Illuminate\Session\DatabaseSessionHandler with PostgreSQL on PHP 8.5

Possible Cause

In DatabaseSessionHandler::getDefaultPayload(), the handler resolves Guard::class to get user_id. During session regeneration after auth, this may create a problematic interaction that triggers segfault in PHP 8.5's pdo_pgsql.

Environment

• Ubuntu 24.04.3 LTS (Docker)
• Architecture: ARM64 (Apple Silicon)
• Laravel Sail

Steps To Reproduce

1. Fresh Laravel 12 + PostgreSQL setup
2. Configure .env: DB_CONNECTION=pgsql SESSION_DRIVER=database
3. Run php artisan session:table && php artisan migrate
4. Create a user
5. Implement login:

if (Auth::attempt($credentials)) {
    session()->regenerate();  // SIGSEGV here
    return redirect('/dashboard');
}
6. Submit login form → PHP crashes with exit code 139

Workaround: Use SESSION_DRIVER=redis

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.

Bug description

Chrome, Firefox, and Safari now support setting the width and height attributes on <source> elements within <picture> elements. You can read more about this in the web.dev article on optimizing CLS. This enhancement allows browsers to better handle layout shifts and varying aspect ratios for responsive images.

How to reproduce

Current Behavior:

Currently, Statamic Responsive Images does not take advantage of this browser capability. As a result, when multiple sources are provided, the appropriate width and height attributes are not applied to <source> elements, which may lead to layout shifts due to varying aspect ratios.

Proposed Enhancement:

Set appropriate width and height attributes on <source> elements based on their aspect ratios.

Benefits:

• Improved Visual Stability: Reduce Cumulative Layout Shift (CLS) by providing browsers with the correct dimensions upfront.
• Better Performance: Modern browsers can allocate the correct space for images during page load.
• Enhanced Responsiveness: Accurately reflect different image aspect ratios for a more consistent user experience.

Environment

Environment
Application Name: local
Laravel Version: 11.41.3
PHP Version: 8.3.16
Composer Version: 2.8.5
Environment: development
Debug Mode: ENABLED
URL: local.test
Maintenance Mode: OFF
Timezone: Europe/Paris
Locale: en

Cache
Config: NOT CACHED
Events: NOT CACHED
Routes: NOT CACHED
Views: NOT CACHED

Drivers
Broadcasting: null
Cache: file
Database: sqlite
Logs: stack / single
Mail: smtp
Queue: redis
Session: file

Statamic
Addons: 3
Sites: 1
Stache Watcher: Disabled (auto)
Static Caching: Disabled
Version: 5.46.1 PRO

Statamic Addons
rias/statamic-redirect: 3.9.3
spatie/statamic-responsive-images: 5.2.1
statamic/seo-pro: 6.5.0

Installation

Fresh statamic/statamic site via CLI

Antlers Parser

regex (default)

Additional details

Leveraging native width and height support on <source> elements can lead to significant improvements in layout stability and overall performance on modern browsers. This enhancement would allow the package to better handle responsive images by using up-to-date browser capabilities.

Thanks for considering this enhancement!

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

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

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

AST, or "abstract syntax tree", creates the expectation of a nested tree structure. In the case of a query string, the AST will only ever be one level deep; as multi nodes are combined into one.

If anyone wants to suggest a better, more clear name: you're welcome.

https://www.algolia.com/doc/libraries/javascript/v5/upgrade/

Calling for help from Docker experts. We need to create the best possible docker-compose.yml file for this project. The application requirements are well defined (we use env vars, Webpack Encore, PHP 7.1, Symfony 4.1, SQLite database, etc.) so it should be possible to create that file.

This polyfill should be implemented in 2 places:

• in symfony/polyfill-intl-grapheme
• in symfony/polyfill-php85, only when the intl extension is loaded (to cover the case of using intl on older PHP versions that won't have this function)

Describe the bug logUnguarded() works provided that $guarded is set explicitly in the model. If it is not set explicitly but globally set, say in AppServiceProvider using Model::unguard(); then nothing is logged.

To Reproduce

1. remove $fillable and $guarded from a model that use uses LogsActivity
2. call Model::unguard(); in AppServiceProvider in the boot method
3. set

    public function getActivitylogOptions() : LogOptions
    {
        return LogOptions::defaults()->logUnguarded();
    }

4. Make some changes to a model and save them

Expected behavior Some changed properties should be logged but none are

Versions (please complete the following information)

• PHP: 8.1
• Database: MySql
• Laravel: 9.38
• Package: larvel-activitylog

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.

Currently, for plural rules, the MessageFormatter polyfills uses the English rules for all locales (it ignores the locale). To be consistent with what we do in symfony/intl (used by symfony/polyfill-intl-icu to implement NumberFormatter and DateFormatter), we should rather fail explicitly here (using the English rules for a different locale would not give the right result anyway)

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

What happened?

I just installed package and complement for AI powered solution. When I create an issue to test it I get :

I had to change the links function like this to make it work for me:

public function links(): array
{
    $rawText = Str::finish($this->rawText, 'ENDLINKS');

    $textLinks = $this->between('LINKS', 'ENDLINKS', $rawText);

    $textLinks = explode(PHP_EOL, $textLinks);

    $links = [];

    foreach ($textLinks as $textLink) {
        $textLink = str_replace('\\', '\\\\', $textLink);
        $textLink = str_replace('\\\\\\', '\\\\', $textLink);

        $decodedLink = json_decode($textLink, true);

        if ($decodedLink !== null) {
            $links[$decodedLink['title']] = $decodedLink['url'];
        }
    }

    return $links;
}

How to reproduce the bug

When I install the package in my project and use it with AI powered

Package Version

1.0

PHP Version

8.2.14

Laravel Version

11.14

Which operating systems does with happen with?

Windows

Notes

No response

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.

While working on the new version of https://github.com/symfony/stimulus-bridge, I noticed some dead code from https://github.com/symfony/webpack-encore/pull/1062:

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

Describe the bug

When you attempt to dump a binary string or anything containing one (like an array or an object), Ray does not dump anything. It just fails silently.

Versions

Ray version (you can see this in "About Ray"): 2.8.1

You can use composer show to get the version numbers of:

• spatie/ray package version: 1.41.2
• spatie/laravel-ray package version (if applicable): 1.37.1

PHP version: 8.2.17 Laravel version (if applicable): 11.18.1

To Reproduce

Add this in your code:

ray(hex2bin('ae0f3d'));

Expected behavior

Ray should display this (like the console when dump is used):

b"®\x0F="

Desktop (please complete the following information):

• OS: MacOs
• Version: 14.6.1

Laravel Version

12.40.2

PHP Version

8.4.15

Database Driver & Version

Postgres 18.1

Description

On Postgres, it appears that the from() / startingValue() calls in a migration file do not appear to do anything.

I have a very simple migration file with following up() method;

        Schema::table('users', function (Blueprint $table) {
            $table->bigIncrements('id')
                ->from(2000)
                ->change();
        });

I do expect something related to alter sequence users_id_seq restart with 2000 as can be seen in; https://github.com/laravel/framework/blob/d3d7c2ec841795db97eb1c1f85da03a3aa6dbeb0/src/Illuminate/Database/Schema/Grammars/PostgresGrammar.php#L276 to be executed on the database server, however, it does not appear to be executed.

The same happens if you use a Schema::create() call, like this for example;

        Schema::create('users', function (Blueprint $table) {
            $table->bigIncrements('id')
                ->from(100);
        });

Based on the previous ::create() statement, the following queries should be sent to the database server;

array:2 [
  0 => "create table "users" ("id" bigserial not null primary key)"
  1 => "alter sequence users_id_seq restart with 100"
]

Yet the ID autoincrement values still start with 1.

If I execute the query alter sequence users_id_seq restart with 100 or alter sequence users_id_seq restart with 2000, manually through my SQL client, it does work.

Steps To Reproduce

1. Create a simple migration like;

        Schema::create('users', function (Blueprint $table) {
            $table->bigIncrements('id')
                ->from(100);
        });

2. Execute it
3. Try to insert a row
4. The primary ID of the table still starts with 1 instead of the expected 100.

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

composer recipes the/package

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.

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

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

Laravel Version

12.39.3

PHP Version

8.2

Database Driver & Version

MYSQL 8.4.3

Description

When I try to import a schema.dump file into a MySQL database over SSL, I can’t provide the --ssl-cert and --ssl-key options, because the import command only supports --ssl-ca. Also, why is the --ssl=off option commented out?

Steps To Reproduce

laravel/framework/src/Illuminate/Database/Schema/MySqlSchemaState.php line 108

protected function connectionString()
    {
        $value = ' --user="${:LARAVEL_LOAD_USER}" --password="${:LARAVEL_LOAD_PASSWORD}"';

        $config = $this->connection->getConfig();

        $value .= $config['unix_socket'] ?? false
            ? ' --socket="${:LARAVEL_LOAD_SOCKET}"'
            : ' --host="${:LARAVEL_LOAD_HOST}" --port="${:LARAVEL_LOAD_PORT}"';

        if (isset($config['options'][\PDO::MYSQL_ATTR_SSL_CA])) {
            $value .= ' --ssl-ca="${:LARAVEL_LOAD_SSL_CA}"';
        }

//        if (isset($config['options'][\PDO::MYSQL_ATTR_SSL_VERIFY_SERVER_CERT]) &&
//            $config['options'][\PDO::MYSQL_ATTR_SSL_VERIFY_SERVER_CERT] === false) {
//            $value .= ' --ssl=off';
//        }

        return $value;
    }

Laravel Version

12.40.2

PHP Version

8.3.25

Database Driver & Version

Microsoft SQL Server 2025 for MacOS Tahoe (26.1) (using Orbstack's Rosetta and Docker platform settings to emulate Intel)

Description

There is a problem with Laravel's Schema Builder when creating JSON columns on SQL Server 2025. Although SQL Server 2025 introduces native JSON datatype support, the ->json() method still creates nvarchar(max) columns instead of using the new native JSON datatype. This results in underutilization of important database features and may negatively impact performance and validation.

Moreover, there is no mechanism to detect the SQL Server version and conditionally use the native JSON datatype only when supported. For older SQL Server versions, the current fallback to nvarchar(max) should of course remain for backwards compatibility.

Steps To Reproduce

1. Create a Laravel migration script defining a JSON column:

Schema::create('example_table', function (Blueprint $table) {
    $table->json('data');
});

2. Run this migration on a SQL Server 2025 database.
3. Check the datatype of the created column in the database.
4. You'll see it's a nvarchar datatype instead of json

NotOrm Package

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

Dashboard showing Operational

Front end showing no issues, but the status seems to still be linked to the incident even if it's closed as fixed

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.

Discussed in https://github.com/spatie/schema-org/discussions/202

Originally posted by indyjonesnl January 10, 2024

$lodgingBusiness = Schema::lodgingBusiness()
  ->openingHours(
    Schema::openingHoursSpecification()
      ->dayOfWeek([Schema::dayOfWeek()::Monday])
      ->opens(new DateTime('09:00:00'))
      ->closes(new DateTime('17:00:00'))
  )
  ->checkinTime(new DateTime('14:00:00'))
  ->checkoutTime(new DateTime('11:00:00'))

This results in the current date + time being included in the output "opens":"2024-01-10T09:00:00+00:00","closes":"2024-01-10T17:00:00+00:00" and ..."checkinTime":"2024-01-10T14:00:00+00:00","checkoutTime":"2024-01-10T11:00:00+00:00"...

While the Schema should contain only the time, formatted as 14:30:00+08:00.

Can someone point me in the right direction?

The friendly-errors-webpack-plugin has a built-in transformer/formatter for when a module is missing. To see the behavior, try (in any JavaScript file):

require('foo'); // foo is not a module that is installed

The error is:

This dependency was not found:

• foo in ./app/Resources/assets/js/app.js To install it, you can run: npm install --save foo

The problem is the npm install part... which should be yarn add to avoid confusion with our users. There is an issue about this: https://github.com/geowarin/friendly-errors-webpack-plugin/issues/33

We are logging activity of several models, some using SoftDeletes trait and others not.

When we set config option 'subject_returns_soft_deleted_models' => true then calling subject() on any model not using SoftDeletes trait throws an exception: Call to undefined method Illuminate\Database\Eloquent\Builder::withTrashed().

Can the subject() method below be made to check first whether the Model uses the SoftDeletes trait? Or perhaps check whether withTrashed() is a defined method on the Model? I tried a few things but couldn't figure out how to get the class of the Model..

https://github.com/spatie/laravel-activitylog/blob/68eb6e65382f94fe86ce5ff6572159f80d024ba9/src/Models/Activity.php#L28-L35

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.

In raising this issue, I confirm the following (please check boxes):

• I have read and understood the contributors guide.
• I have checked the pull requests tab for existing solutions/implementations to my issue/suggestion.
• I have checked that the bug-fix I am reporting can be replicated.

Description of the problem

One of the important reasons for using hyperscript is that it handles escaping of characters which are reserved in html. Which prevents XSS.

spatie/html-element does not do this. Mainly because it uses strings as intermediate format instead of an object representation of the DOM.