Put each entity type table into a details element on admin/config/regional/content-language

Made new MRs for 9.5 and 10.1:

- 9.5 https://git.drupalcode.org/project/drupal/-/merge_requests/3846
- 10.1 https://git.drupalcode.org/project/drupal/-/merge_requests/3848

I don't understand the test failure:

WebDriver\Exception\ElementNotVisible: element not interactable
  (Session info: headless chrome=106.0.5249.103)
  (Driver info: chromedriver=106.0.5249.61 (511755355844955cd3e264779baf0dd38212a4d0-refs/branch-heads/5249@{#569}),platform=Linux 4.9.0-0.bpo.6-amd64 x86_64)

at /var/www/html/core/modules/content_translation/tests/src/FunctionalJavascript/ContentTranslationContextualLinksTest.php:51

The line in question is:

    $this->submitForm($edit, 'Save configuration');

The 'Save configuration' button is always visible -- it's not inside a details element.

@joachim

Then the element is not in the viewport. Just scroll to the submit button :D

But in both tests is the same error:

POST Content-Length of 8392180 bytes exceeds the limit of 8388608 bytes in Unknown on line 0

https://dispatcher.drupalci.org/job/drupal_patches/179196/artifact/jenkins-drupal_patches-179196/artifacts/run_tests.javascript/test.apache.error.log/*view*/
https://dispatcher.drupalci.org/job/drupal_patches/179184/artifact/jenkins-drupal_patches-179184/artifacts/run_tests.javascript/test.apache.error.log/*view*/

Thanks for the pointers for the test failure.

It should actually be fixed by 📌 ContentTranslationContextualLinksTest should use API calls to set up translation Fixed -- we should not be submitting this form in a test's setUp()!

I've removed the #collapsible property -- that seems to be a leftover from an earlier version of this MR when it was a fieldset element.

Postponed on that other issue, which will fix the test failure.

Hopefully now that other issue is fixed, this will pass -- made a new MR on 11.x.

Very simple MR and yet very impactful until a better solution is found. Tests failures are apparently fixed by the other issue.

I'm triaging RTBC issues → . I read the IS and the comments. I didn't find any unanswered questions or other work to do.

This looks like a good improvement to me

Yes, I agree! Thanks for working on this.

Leaving at RTBC.

Saving issue credits.

There's some MR chaos here. Let's please have one canonical merge request that is mergeable and passing tests, for 11.x, and close the others. We generally cherry-pick backports in any case, and this is a minor-only change so would only target 11.x/10.2.x which are currently identical.

If you are unable to close the MRs yourself, please document which MRs should be closed in an issue comment and the IS, and a committer can then close them for you.

The MRs all say in their branch name which core branch they target (except the oldest one, before this issue dragged on. The canonical one has 10.2 in the name.

I don't want to close the MRs because the diffs from them are useful as patches - again, because this issue has dragged on.

Apologies, but I'm going to have to overrule this here (it's within the release management purview to make final decisions about issue workflow and process).

As a policy, we need to have only one canonical change set listed in the IS (except for certain kinds of issues like cross-branch dependency updates). This issue is also a minor-only change, and therefore not backportable to patch-supported or security versions, so the 10.1 and 9.5 branches will not be used for core.

I am fairly sure MRs can be accessed and their diffs used for composer-patches even if they are closed. Or, a file could also be attached to an issue comment and hidden, and would still be available.

So, I'm closing all the old/broken/failing MRs.

xjm → closed merge request !3846

xjm → closed merge request !3848

xjm → closed merge request !594

Waiting for the pipeline to finish, but this looks committable to me. Great UX improvement.

Committed to 11.x and cherry-picked to 10.2.x. Thanks!

xjm → closed merge request !4614

Automatically closed - issue fixed for 2 weeks with no activity.