reproduction for fix-match-loading edge-cases#7713
Conversation
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
|
View your CI Pipeline Execution ↗ for commit 4d218c3
☁️ Nx Cloud last updated this comment at |
| * | ||
| * This test navigates through the public router API with object-form | ||
| * viewTransition options. The mocked browser reports no typed-transition | ||
| * support, but still provides document.startViewTransition, so navigation should |
There was a problem hiding this comment.
does this behavior exist in actual browsers?
There was a problem hiding this comment.
it seems to be true that the case exists in actual browsers, but it may be a somewhat narrow band
Document.startViewTransition: Chrome 111, Safari 18, Firefox 144:active-view-transition-type(): Chrome 125, Safari 18.2, Firefox 147
i'm not 100% convinced it's worth jumping through hoops to support this. I'll leave that up to you
| * loading can behave differently from other route failures. | ||
| * | ||
| * This test uses a real router navigation to a lazy route whose chunk rejects. | ||
| * The match reaches error state with the chunk error, but the route's onError |
There was a problem hiding this comment.
isn't it weird if onError isn't called in the specific case of "failed to load a chunk", but called in all other error cases?
| import { createTestRouter } from './routerTestUtils' | ||
|
|
||
| /** | ||
| * Issue 8: route headers are response behavior and should not be dropped just |
There was a problem hiding this comment.
should a head throwing on the server be swallowed in general? or is the whole response than a 500 with no data?
There was a problem hiding this comment.
I think this issue here is about the inconsistency: if we don't fail the request on head() failure, we shouldn't drop the headers() on head() failure.
But I feel like a fail is a fail, and it should 500 instead of delivering partial content.
2 participants
No description provided.