Tags: microsoft/FluidFramework
Tags
build: release notes and changelogs for 2.110.0 (#27586) ## Description Generate release notes and per-package changelogs for the upcoming **2.110.0** minor release. Per [minor release prep](.claude/skills/fluid-release/references/minor-release-prep.md) Step 3. ### Commits 1. `build: add release notes for 2.110.0` — `RELEASE_NOTES/2.110.0.md` (generated by `flub generate releaseNotes -g client -t minor`) 2. `build: generate changelogs for 2.110.0` — 11 changesets consumed, 167 `CHANGELOG.md` files updated across the client release group (generated by `flub generate changelog -g client`) ### Release prep order This PR is **#3 of 4** for the 2.110.0 release prep. Step 2 (compat-gen) was a no-op (27 days since last gen, below the 33-day threshold). Must merge before the version bump PR (#4). ### Related PRs - #27585 (Step 1: tag asserts) --- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> --------- Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Update test versions (#27519) ## Description pnpm run --filter=@fluid-private/test-version-utils update-compat-versions
fix(tree): batch uniform-shaped nodes into multi-node UniformChunks (#… …27181) ## Summary This is the final PR to enable multi-node chunks in chunked Forest. This PR should not be merged until after [support attach/detach edits on multi-node chunks](#27153), [track chunkIndex on field entry so getField finds parent](#27084) `chunkRange` was evaluating `maxTopLevelLength` as `ceil(nodesPerTopLevelNode / uniformChunkNodeCount)`, which is the inverse of the intended ratio. For a leaf shape with `nodesPerTopLevel = 1` (eg. plain text) with `uniformChunkNodeCount = 4`, this produced `1`, so each `UniformChunk` held a single node and batching never happened. The fix flips the division and clamps to `1`: <pre> const maxTopLevelLength = Math.max(1, chunkCompressor.policy.uniformChunkNodeCount / nodesPerTopLevelNode); </pre> A 10-element number field with `uniformChunkNodeCount = 4` now produces chunks of length `[4, 4, 2]` instead of ten singletons. #### Tests | File | Changes | | ---- | ---- | | chunkTree.spec.ts | Adds `uniformChunks` describe block. "batches uniform shaped nodes into chunks of uniformChunkNodeCount" - direct `chunkRange` over a 10-node field, asserts [4, 4, 2]. Uses a custom chunk policy. Added a single regression test for calling splitFieldAtIndex after chunkRange with a custom chunkPolicy that tests batching is done correctly, and that the bisection splits the uniformChunk with the intended behavior| #### Notes - ForestSummarizer has a single test that assumes each node gets it own chunk. Since this is no longer the case the test has been updated to assert 1 uniform chunk instead of 2 - Any operation that requires `enterNode` to be called on a uniformChunk with `topLevelLength = N`will cause the chunk to explode into N basicChunks. There is a TODO to not require this, but for now this is the intended behavior. ## Reviewer Guidance The review process is outlined on [this wiki page](https://github.com/microsoft/FluidFramework/wiki/PR-Guidelines#guidelines).
build: generate release notes and changelogs for 2.101.0 (#27301) Generated by pnpm flub generate releaseNotes -g client -t minor and pnpm flub generate changelog -g client. Adds RELEASE_NOTES/2.101.0.md, deletes 8 consumed changesets, updates 175 package CHANGELOG.md files. --------- Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Use shared config throttler (#27201) ## Description Use shared configThrottler instead of build wheel from scratch ## Breaking Changes N/A Co-authored-by: Xin Zhang <zhangxin@microsoft.com> Co-authored-by: Copilot <copilot@github.com>
PreviousNext