revert: drop clustering dim-rename layer via tsam-xarray DimNames (#739) - #746
Conversation
|
Warning Review limit reached
Next review available in: 5 minutes Enable usage-based reviews in Billing to review now. Otherwise, wait until the next included review is available. How can I continue?After more reviews become available, a review can be triggered using the To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews. How do review limits work?CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability. For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window. Please refer docs for additional details. Review details⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (4)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
Retarget the pending release from 8.0.0 to 7.3.0. The major came solely from the BREAKING CHANGE footer on #739, which was reverted in #746. Release-please does not pair a revert with its original, so the footer still counted. The behaviour it describes is not in the released code: files written by 7.2.3 load and expand unchanged on main, and since #757 files written by 6.1.0 and 6.2.1 load as well. The breaking notice is dropped and #739 is listed under Reverts instead of Code Refactoring, since it nets to zero in this release. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
8.0.0 shipped with a breaking-change notice describing a clustering format change that is not in the release. #739 introduced it and #746 reverted it before release, but release-please does not pair a revert with its original, so the BREAKING CHANGE footer both drove the major bump and rendered into the notes. Replaces the notice with a note explaining why the major carries no breaking change, and moves #739 from Code Refactoring to Reverts. Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Why revert
#739 was labeled
refactor:but carries a breaking change for persisted clustering artifacts: clustering JSON/netCDF saved by any flixopt before it no longer loads (_periodslice dim vs liveperiod). Under release-please that breakage would ship silently in a v7.x minor.Policy going forward: main stays fully v7-compatible; breaking changes collect on the
v8integration branch. #739's change lives on there (thev8branch contains it) and will ship with v8.0.0, where the artifact break is covered by the major version and the migration guide.Scope
TestLegacyClusteringBackwardCompatsuite) stays in place.tsam_xarray >= 0.6.1minimum.Verification
tests/test_clustering+tests/test_math/test_clustering.py: 275 passed, 0 failed.Merging main into
v8after this lands will try to apply the revert there too. The next sync must revert this revert onv8(or resolve the merge keeping v8's side) so #739's change survives where it belongs.🤖 Generated with Claude Code