[core] Support `@mui/material@6` peer dependency
Fixes https://github.com/mui/mui-x/issues/14055. Fixes https://github.com/mui/mui-x/issues/14369.
TODO:
- [x] Automate
@mui/materialversion check - [x] Test a package install with
@mui/material@next - [x] Test all the changes with
@mui/material@5 - [x] Create a script that will set
@mui/materialversion tonext - [x] Add
test_typesjob tomaterial-ui-v6workflow - [ ] Make the
material-ui-v6workflow pass - [ ] Schedule
material-ui-v6workflow? - [x] Use Argos for
material-ui-v6workflow? Is it possible to have 2 Argos checks on the PR? Edit: the default workflow sends screenshots to Argos. Thematerial-ui-v6workflow also sends screenshots to Argos effectively overriding the ones from the default workflow. I think it's OK.
Deploy preview: https://deploy-preview-14142--material-ui-x.netlify.app/
Generated by :no_entry_sign: dangerJS against e64adb1ee9bb4faf4cc7f6bffdb3a3721766bf0d
This pull request has conflicts, please resolve those before we can evaluate the pull request.
This pull request has conflicts, please resolve those before we can evaluate the pull request.
This pull request has conflicts, please resolve those before we can evaluate the pull request.
This pull request has conflicts, please resolve those before we can evaluate the pull request.
This pull request has conflicts, please resolve those before we can evaluate the pull request.
This pull request has conflicts, please resolve those before we can evaluate the pull request.
This pull request has conflicts, please resolve those before we can evaluate the pull request.
The errors in next tests are the same as I saw when doing the opposite—using v6 material, but v5 @mui/utils and @mui/system.
It looks like there might be an incompatibility in RTL resolving between the majors. 🙈
@mui/core are you aware of anything that could cause sporadic useRtl hook behavior when using @mui/system@5 + @mui/material@6 or the other way around (@mui/system@6 and @mui/material@5)?
This pull request has conflicts, please resolve those before we can evaluate the pull request.
The errors in
nexttests are the same as I saw when doing the opposite—using v6 material, but v5@mui/utilsand@mui/system. It looks like there might be an incompatibility inRTLresolving between the majors. 🙈 @mui/core are you aware of anything that could cause sporadicuseRtlhook behavior when using@mui/system@5+@mui/material@6or the other way around (@mui/system@6and@mui/material@5)?
@LukasTy I assume the issue is solved (all checks are green)
I assume the issue is solved (all checks are green)
@aarongarciah not sure why GH is reacting like this, it showed the failed steps previously... 🤷 The issue is not solved: https://app.circleci.com/pipelines/github/mui/mui-x/64664/workflows/36635495-f2c1-4713-9bd6-128133c71cb6
not sure why GH is reacting like this, it showed the failed steps previously... 🤷
I've added a separate workflow in CircleCI for running tests against Material UI v6.
It's triggered manually for now, similarly to https://github.com/mui/mui-x/blob/master/test/README.md#next-version (the workflow name is material-ui-v6). I'll add this to that readme.
@LukasTy FYI – the material-ui-v6 workflow is not triggered automatically, you can trigger it manually following the instructions: https://github.com/cherniavskii/mui-x/blob/material-ui-v6/test/README.md#ci-1
This pull request has conflicts, please resolve those before we can evaluate the pull request.
This pull request has conflicts, please resolve those before we can evaluate the pull request.
This pull request has conflicts, please resolve those before we can evaluate the pull request.