🇫🇷France @pbonnefoi

It seems that this is not working on the latest version. I'm investigating.

pbonnefoi → made their first commit to this issue’s fork.

Great work @nils.destoop !

Last patch from #155 works great on 10.2.5

I'm having the same issue

pbonnefoi → created an issue.

Patch #35 for 1.11 applied on d10.2.3. as well and fixes my problem.

I think accordion should be removed from the module and be in another sub mobule. It's also consufing because "Detail" field group is also like an accordion.

Glad I could help !

Ok good to know, thank you very much for the insights !

I have a similar problem trying to fix a unit test on a contrib module, but adding $this->getSession()->reset(); did not work. Do you add it right after drupalLogout() ?

An improvement on the error message would be very nice as well because it does not make much sense.

The unit test causing problem is from the redirect_after_login contrib module.

Well, you might be right. I thought about it but didn't know how to proceed.

Hello, the module does not handle custom entities with no bundles yet, but feel free to contribute if you have some spare time :-)

Indeed it was accessible to anonymous user. I changed it and it's kinda working but it's odd for me that the bearer is not checked prior to the content access permission and besides there is no logging that helps you really understand why it's not accessible.
I have the problem on another project where we have a custom rest route with permission only to authenticate users and it's not working. Maybe it's du to accessCheck on query not there (as now it's mandatory since D10).

@johan-den-hollander I opened an MR with the patch from #2765895 (with little improvements) I have other ideas, but need to find time to implement it.

I made this feature as an option with the following settings :

You can set flush_redis_on_drupal_flush_cache to TRUE to flush Redis cache when clearing all Drupal cache.

$settings['flush_redis_on_drupal_flush_cache'] = TRUE;

You can set flushdb and flushdb_async to TRUE to use the Redis native flushDB when clearing Drupal cache.

$settings['redis_flushdb'] = TRUE;
$settings['redis_flushdb_async'] = TRUE;

You need to set a redis base properly to use this feature. By default, Redis gets cleared getting all keys using cache_prefix. This is not recommended on production as it could be an expensive task to clear all the keys.

pbonnefoi → made their first commit to this issue’s fork.

pbonnefoi → made their first commit to this issue’s fork.

pbonnefoi → created an issue.

pbonnefoi → created an issue.

Thanks, it was merged and released in both 8.x-2.3 a,d 3.0.0. Thanks @chipway

Thanks @Soubi for the report, duplicate of #3005314. I will credit you as well for the patch.

Available since the release 8.x-2.1

pbonnefoi → made their first commit to this issue’s fork.

pbonnefoi → made their first commit to this issue’s fork.

Great ! Thank you !

All PHPCS and PHPStan error fixed

No matter what I do, I still have the message to rebase. I followed this documentation here => https://www.drupal.org/docs/develop/git/using-gitlab-to-contribute-to-dr... → and was able to rebase, but still not working, I tried once again with accepting all commits from 3.0.x prior to mine but it's not working...

Well, in my opinion, since the module is using js and css, eslint must be enabled. Adding gitlab-ci to modules is the new norm. Helps improve the code quality and best practices overall. We will open other tickets to fix code issues.

pbonnefoi → created an issue.

@VladimirAus good to merge now after rebase :-)

Add the .gitlab-ci.yml to run PHPCS, PHPStan and PHPUnit and improve future code quality of the module.

Well I added it because it was necessary to see if all errors were fixed. What do you suggest I should do ?

I think the 3.0.x needs to stay as it is the new version standard (And I've done some work to update the code to the latest standards). A release in 3.0.0 would be nice and let the branch 8.x be only the one working on Drupal 8 and 9.

All PHPCS and PHPStan errors are fixed and PHPUnit up and running. Needs code review and merge I guess (with a new version)

pbonnefoi → created an issue.

Gitlab-ci is up and running. Opening a new issue to fix all PHPCS and PHPStan issues

pbonnefoi → created an issue.

Duplicate of 3289274 📌 Automated Drupal 10 compatibility fixes Fixed

It's actually ready to merge here => https://git.drupalcode.org/project/redirect_after_login/-/merge_requests...

Just waiting for the maintainer to merge it. Here is the patch to apply if you need it => https://git.drupalcode.org/project/redirect_after_login/-/merge_requests...

pbonnefoi → created an issue.

Thanks @perrypierre for the patch. I mentored @kibbas on doing MR with the new gitlab system, now waiting for the maintainers to merge it.

@pmagunia, I don't recommend to set it to FALSE as this would imply some Authorization & access control security issues.

I added the gitlab-ci file and updated the code to the last PHPCS standard. I tried to fix the PHPStan Errors, but I don't really know what's wrong.

79 Parameter $queue of function auto_entityqueue_add_entity_to_queue()
has invalid type Drupal\entityqueue\EntityQueueInterface.
90 Call to static method loadMultiple() on an unknown class
Drupal\entityqueue\Entity\EntitySubqueue.
💡 Learn more at https://phpstan.org/user-guide/discovering-symbols
128 Call to static method loadMultipleByTargetType() on an unknown class
Drupal\entityqueue\Entity\EntityQueue.
💡 Learn more at https://phpstan.org/user-guide/discovering-symbols

pbonnefoi → created an issue.

pbonnefoi → made their first commit to this issue’s fork.

ckrina → credited pbonnefoi → .

@saschaeggi I saw your commit but that did not solve the translation issue. In my last commit I changed to use the routing instead of the title to create the class. It's less likely to change than a title I think so more robust (despite having not so beautiful classes). I also made changes on Create and Content navigation to make it consistent.
Works fine on my version (But still having sprite issues for some reasons, but I think that is another problem).

I might find a better solution later with creating a custom manipulator. It's funny I saw an interesting solution on my own question asked on Stackexchange posted 7 years ago :-D (https://drupal.stackexchange.com/questions/201903/programmatically-get-m...)

Anyway, let me know what you think and if I can help on other things.

@NicolasGraph I tried the patch but some icons are still missing (using french language)

Here is a patch we made with our DevOps in order to avoid this problem.

Also, I should mention that one problem has not been addressed yet. If all redis instances are down, then the site is down. There is a solution for a failover as mentioned here => https://www.drupal.org/project/redis/issues/3102739 ✨ Allow failing over to database or other backends if Redis is down RTBC
But it works only if you have one instance configured. We should manage to have a failover also in a multi instance configuration. I don't have the solution yet.

The #28 patch works great with a single Redis instance, but there one use case that is not covered yet.

I have a Redis with Sentinel Master/Slave and therefore, my settings is having something like that :

$settings['redis.connection']['host']      = ['XXX.XXX.X.X:XXXX', 'YYYY.YYY.Y.Y:XXXX];

When one of the redis is down, caching is still working thanks to the patch here => https://www.drupal.org/project/redis/issues/3163436 🐛 Connection refuse in PhpRedis when sentinel/redis are down RTBC

But when both are down, then the failover is not working. I'll try to find a solution and post it if so.

Can someone validate the merge request here => https://git.drupalcode.org/project/auto_entityqueue/-/merge_requests/1 in oder to update it properly with composer ? Thanks :-)