-
Notifications
You must be signed in to change notification settings - Fork 1.3k
fix useRangeCalendar behavior while interacting outside #5721
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
I don't know why but it seems like there were a few test case failed on commits before my. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for the PR, and the detailed testing instructions!
First, you'll need to sign the Adobe CLA to get that build step to pass.
And it looks like the tests that are failing are correctly failing due to the updated behavior (i.e. clicking outside calendar commits selection
). We'll need to update those to match the new expected behavior.
If any tests were failing before your changes, they were probably related to a node version mismatch and can be ignored (they'll pass in CI).
I've signed Adobe CLA and made CI running. Did "updated behavior" mean the code I submitted? If so, could you point me which test cases should I take care of? It's currently too many tests failed on my machine (node v18.18.2 on commit ee51290). I'm not sure these test cases were "correctly failing" or caused by my environment. |
@QzCurious Could you try node version |
It works on The relevant failed cases are:
And I've changed those to match new behavior:
|
By the way, On my mac book, there are other failed tests. Seems like some were caused by encoding. And some might be platform specific. The output text here |
GET_BUILD |
Build successful! 🎉 |
Hi, just wanted to know how it goes. Let me know if there's anything I can help with. |
@QzCurious apologies, we've got a bit side tracked with last release, I'll see if we can get this into an upcoming testing session for the team to evaluate the new behavior. |
GET_BUILD |
Build successful! 🎉 |
## API Changes
unknown top level export { type: 'any' } |
@QzCurious apologies about the delay, but the team went through the behavioral changes in a testing session today. Overall, we liked the update in behavior that allows the user to now scroll through multiple calendars and finalize a range selection. However, we weren't sure about completely canceling the range selection upon clicking outside the calendar, namely because of a flow where the user might assume that their selection has been made and then they click on an accompanying TimeField to continue configuring their date as shown here: Screen.Recording.2024-02-26.at.4.23.58.PM.movwith the new changes that would get canceled: Screen.Recording.2024-02-26.at.4.24.54.PM.mov |
I understand there should have some effort for validating all possible scenarios. And im gald you found a "behavioral bug" for my change. I'd love to ask if you team would have a plan to sort it all and implement it (because it actually involve some decisions instead of just a bug). Or I could still submit some progress, if any, for you to validate it. By the wayFor scroll through calendars, first recording is wrong. You should validating it on mobile interface (you can refer to the videos on #5721 (comment)). |
From the team's discussion yesterday, we'd be happy to accept a change isolated to fixing the mobile scrolling range selection but it was hard to tell from a glance if that would be easy to pull out from the changes made in this PR. Would that be possible?
As for the first recording, that is intentionally showing the desktop behavioral change rather than the mobile behavior. We were concerned that the changes may be confusing to a user using the RSP DateRangePicker because the flow shown in the first video here used to allow a user to confirm a date range selection when they click into the neighboring TimeField but that wouldn't happen after these changes in the PR. |
I see. Thanks for explaining it! Let me know if you want keep these PR and issues open, or I will close this PR and related issues. And open an issue for "mobile scrolling range selection". |
Feel free to close them and open a new issue! |
Close. Summery
|
Closes #5703, #5709
This is a proposal to fix
useRangeCalendar
behavior by directly showcasing what was wrong (IMO).Most problem were state on #5703. But I just can't state the issue any better. This PR would hopefully make things more clear.
✅ Pull Request Checklist:
📝 Test Instructions:
The videos is recorded with existing story.
The main difference is instead of commit selection, I canceled it. And canceling means that
onChange
would not get called.Screen.Recording.2024-01-20.at.12.23.54.AM.mov
Screen.Recording.2024-01-20.at.12.25.01.AM.mov
Screen.Recording.2024-01-20.at.12.41.27.AM.mov
Screen.Recording.2024-01-20.at.12.42.41.AM.mov
Screen.Recording.2024-01-20.at.1.03.47.AM.mov
Screen.Recording.2024-01-20.at.1.02.52.AM.mov
Screen.Recording.2024-01-20.at.12.46.00.AM.mov
Screen.Recording.2024-01-20.at.12.44.43.AM.mov
Screen.Recording.2024-01-20.at.12.51.30.AM.mov
Screen.Recording.2024-01-20.at.12.50.52.AM.mov
Screen.Recording.2024-01-20.at.12.54.00.AM.mov
Screen.Recording.2024-01-20.at.12.54.50.AM.mov
🧢 Your Project:
https://github.com/QzCurious/react-spectrum