| Comment | File | Size | Author |
|---|---|---|---|
| #8 | 2998601-8.patch | 6.34 KB | gabesullice |
| #8 | interdiff.txt | 952 bytes | gabesullice |
| #6 | 2998601-6.patch | 5.41 KB | gabesullice |
| #6 | interdiff.txt | 2.99 KB | gabesullice |
| #4 | 2998601-4.patch | 2.43 KB | gabesullice |
| Comment | File | Size | Author |
|---|---|---|---|
| #8 | 2998601-8.patch | 6.34 KB | gabesullice |
| #8 | interdiff.txt | 952 bytes | gabesullice |
| #6 | 2998601-6.patch | 5.41 KB | gabesullice |
| #6 | interdiff.txt | 2.99 KB | gabesullice |
| #4 | 2998601-4.patch | 2.43 KB | gabesullice |
Comments
Comment #2
gabesulliceComment #3
gabesulliceComment #4
gabesulliceHrm.
Comment #5
wim leersThis was introduced in #2984647: Dangling entity references in entity reference field with multiple possible target bundles: results in exception/500 response. Why didn't it cause tests to fail back then?
Why all these changes?
Comment #6
gabesullice1. It looks like it was removed in #2984886: Trigger route rebuild when new bundles/fields are added/removed, which also added a regression test. So, the better question is, why didn't that fail tests? I think the answer is because it doesn't fail on PHP7 and #2984886 didn't have a PHP5.5 test run against it.
2. Because the (
$var = call())->save();pattern was failing PHPLintConstants can't be arrays on PHP 5.5...
Comment #7
gabesulliceComment #8
gabesulliceComment #9
gabesulliceAll tests pass. Committing because this is critical (it's failing other patches for no good reason).
Comment #11
gabesullice