Repository navigation
Release v0.6.6 — correctness fixes, docs alignment, evaluation evidence - #27
Merged
Merged
Conversation
Drop production_suite_test, soak_test, real_world_latency_spikes, quality_and_readheavy and readratio_isolated from the packaged crate; they stay in-repo and remain runnable locally via autodiscovery. Six teaching examples (basic, concurrent, all_scenarios, memory_baseline, drain_latency, hitrate_16384) still ship. Shrinks the crates.io payload from eleven examples (~90 KiB) to six (~20 KiB).
Removing their [[example]] entries dropped the required-features = ["std"] guard, so autodiscovery built them under --no-default-features and the no_std CI test failed (they use the std-only concurrent types). Set autoexamples = false so only the six explicitly listed examples are targets; the excluded harnesses stay in-repo as reference. Correct the CHANGELOG note accordingly.
cargo tolerates an excluded [[example]] whose source is in exclude: it drops the target from the package (with an "ignoring example" warning) yet still builds it locally. So keep all five harnesses as declared examples with required-features = ["std"] (runnable via cargo run --example, skipped by the no_std build) and just add their files to exclude. Drops autoexamples=false; no orphan files, nothing deleted.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
v0.6.6 — correctness fixes, docs alignment, evaluation evidence
Backward-compatible (patch-level) correctness release over published
0.6.5.This is a bugfix/validation patch, not a performance release — some concurrent
workloads are measurably slower (see trade-offs below).
Fixed
survive concurrent manual, automatic and sharded resize.
indexing on every sharded CRUD/rehash path. Above 16,384 buckets/shard the ranges
previously overlapped; the 1M-key probe now retains 967,051 vs 821,839 residents
at 32,768 buckets/shard.
same event being delivered twice during overlapping wraparound drains; removes the
silent 24-bit bucket-index truncation.
Documentation & evaluation
len()/iteration, insertion-epoch TTL,inline/slab allocation, actual read locks, eviction formula. README/mdBook examples
compile in the rustdoc suite.
evaluation/workspace running registry0.6.5vs the localcandidate in one executable with pinned competitors (LRU/QuickCache/Moka).
package (still in-repo, runnable locally); six teaching examples still ship.
Measured trade-offs (disclosed, accepted)
usizeatomics: +48 KiB per Concurrent map(+768 KiB over 16 shards). Bounded try operations, not a formally lock-free FIFO.
0.6.5exactly.sharded u32 RSS ~61.7 vs ~48 B/resident. No universal speed/memory win claimed.
Validation
Local checks green: stable/nightly tests, no_std, fmt/clippy, strict docs, Loom, Miri,
MSRV, eight embedded targets, both QEMU targets,
cargo publish --dry-run. Seeevaluation/FINAL_VALIDATION.mdandevaluation/PROGRESS.md.