Repository navigation
Releases
v0.0.24 — reusable native read snapshots
Reusable native read snapshots
Native ADBC connections can retain one committed managed catalog generation before metadata discovery and reuse it across sequential table reads. This supplies the driver consistency boundary needed for multi-table full-refresh extraction.
- Acquire/release with
adbc.redis.read_snapshot_action; inspectadbc.redis.read_snapshot_status. - Set a fixed lifetime through
adbc.redis.read_snapshot_retention_ms(120000–3600000 milliseconds; default 900000). - Retain exact managed values, schemas and view definitions across supported concurrent DML/DDL, TRUNCATE/DROP and compaction.
- Fail explicitly on expiry/lost authority, including catalog-only or nonmatching-filter GetObjects discovery. Never fall back to live reads.
- Preserve native authentication and SQL grant/ACL checks. Parent release permits active readers to finish under their original view.
The standard native and installed-wheel validation suites now exercise the two-table shared-snapshot contract. Local complete candidate-wheel validation passed 567 tests with 8 existing unsupported-capability skips and no failures.
Storage protocol 17 is unchanged. Existing fresh-storage and legacy-format rejection rules remain in effect. This is not a multi-statement write transaction, durable sync checkpoint, cross-connection/process attachment or independent-shard snapshot.
Connector integration
The current Airbyte connector still rejects read_consistency=shared_snapshot. Pin this released driver, implement the documented native connection lifecycle before discovery/schema validation, and validate the connector before enabling the setting. A driver upgrade alone does not enable connector support. Flight SQL shared snapshots remain unsupported.
API, Python example and exact Airbyte integration steps
v0.0.23
Redis ADBC v0.0.23 adds atomic statement snapshots for managed tables and explicit retained operations that another process can inspect, resume or abort.
Changes
- Enable managed-table statement snapshots with immutable generations, reader pins, atomic publication and bounded background compaction. (#284)
- Require complete typed input capture before publishing managed HASH imports. Application-owned source HASHes and indexes remain unchanged. (#332)
- Batch manifest proofs and compaction reads, reuse dirty-table descriptors, reduce selective-read allocations, and release HASH-copy inputs promptly. (#346, #347, #348, #349, #351)
- Add recovery-proof diagnostics and preserve failed test-fixture evidence. (#350)
- Add explicit process-resumable operations for evaluated SQL writes, consumed Arrow input and captured managed imports, with fenced ownership, bounded retention and durable outcomes. Resume does not reevaluate SQL or consume application input again. (#355)
- Consolidate integration harness inventories, reuse aggregate expression traversal, and clarify historical compatibility and benchmark evidence. (#333, #334, #352, #353, #354)
Compatibility and recovery
Storage protocol 17 rejects earlier driver catalogs. There is no in-place upgrade or migration path. Use a fresh isolated driver database, or an explicitly supported managed import alongside application-owned data without legacy driver metadata. Preserve a verified backup and the previous deployment before changing versions; replacing the binary does not roll back data or metadata changes.
Snapshots are always enabled for managed tables. Multi-statement transactions remain unsupported. Retained operations require complete input preparation; an interrupted, uncertain native Search preparation phase can make an operation abort-only. Retention defaults to one hour, with documented limits. See the snapshot and operation contract.
Redis Search/Query Engine and RESP3 are required. Redis Flex is rejected. No full-scale performance-parity claim is made for this release.
Distribution and validation
Source: 08d75d50888374b805a5c89f5bb94a46cf6fe8a9.
The merged PR's test workflow and packaging workflow passed their executed checks. PR packaging skipped DuckDB and asset upload as designed; tagged release validation is separate.
Tagged release validation succeeded across all twenty validation jobs, followed by successful asset upload. It covers native builds, ten Linux distribution checks, DuckDB extensions and installed-wheel clients. The initial Ubuntu 22.04 amd64 job stopped before driver tests because its OS package mirror returned HTTP 404; a failed-job-only retry passed on a fresh runner with the same tagged source and test settings. No Go-test skip override was used.
All 33 downloaded assets match their GitHub SHA-256 digests; all 16 published checksum files match their payloads. The three wheels also passed the repository validator for platform set, version, metadata and native libraries. Linux arm64 installed-wheel ADBC runs reported 565 passed and 8 skipped, with all 16 required native partition tests passing; the tagged job logs retain environment-specific skips.
Automatic PyPI publication succeeded. All three PyPI 0.0.23 wheels have the same SHA-256 hashes as their GitHub release counterparts. Distribution is complete.
Install Python:
python -m pip install 'adbc-driver-redis[dbapi]==0.0.23'
python -m adbc_driver_redis doctor --offlinev0.0.22
What's Changed
- docs: align README support labels with completed contracts by @alberttwong in #307
- Verify Software suite runner after TLS migration by @alberttwong in #308
- Store Arrow Map and tagged Union values by @alberttwong in #309
- Enforce table SELECT grants through dedicated Redis Software ACLs by @alberttwong in #310
- Extend bounded Arrow streaming and ADBC interoperability by @alberttwong in #311
- Run Superset integration independently of driver suites by @alberttwong in #312
- Extend Arrow C Device layouts and add Python staging adapter by @alberttwong in #313
- Use latest Redis Software with verified QPF and document Flex 8.8 results by @alberttwong in #315
- Implement bounded Substrait 0.108.0 plan execution by @alberttwong in #316
- Fix Substrait C validation against installed wheels by @alberttwong in #317
- Move Arrow and ADBC limits into dedicated documentation by @alberttwong in #318
- docs: investigate Redis Flex numeric fields and cursor-free reads by @alberttwong in #319
- Add opt-out product usage feedback and deployment analytics by @alberttwong in #244
- Reduce duplicate PR CI work and unblock Linux wheel validation by @alberttwong in #321
- Preserve materialized-view generations for active readers by @alberttwong in #289
- Add actionable driver analytics and compact workload details by @alberttwong in #320
- Add experimental Redis Flex stopgap with bounded range reads by @alberttwong in #323
- Extend Software Flex replay/rekey support and improve bounded reads by @alberttwong in #331
Full Changelog: v0.0.21...v0.0.22
v0.0.21
Redis ADBC v0.0.21 enables bounded prefetch for eligible wide reads and adds experimental CUDA predicate execution and Arrow C Device interchange.
Changes
- Enable prefetch by default for managed-table streams with at least 64 needed columns, positional HASH projection and multiple discovery pages. Ordinary and public partition reads share this path. Each reader owns at most its current reply page plus one lookahead page; variable-width values determine memory cost. The build tag
scanprefetchoffrestores synchronous fetching. #297, #299. - Add opt-in experimental CUDA integer predicates and Arrow C Device import/export, binding and result streams. CPU execution remains the default. CUDA requires Linux cgo, compatible NVIDIA hardware and CUDA runtime/NVRTC libraries; supported query shapes are deliberately limited. Functional checks on NVIDIA A10G do not establish a query speedup. #290.
- Add exact integer SIMD experiments and profiling. Automatic SIMD SQL batching remains disabled because end-to-end measurements did not establish a benefit. #298.
- Upgrade Go source builds to Go 1.27.3 and macOS native packages to macOS 15+. Add typed-export feasibility/design evidence; a server export command is not implemented.
Compatibility
Python requires 3.12+. Native distributions target Linux amd64/arm64 with glibc 2.28+ and macOS arm64 on macOS 15+. The macOS minimum increases from 12 to 15 in this release. Redis Search/Query Engine and RESP3 are required. Transactions and Redis Flex remain unsupported. Prefetch adds no snapshot guarantee or strict byte quota; concurrent readers each have their own lookahead page.
Source and validation
Source: b1d1bf005780516a17ceaf46ed49840506c77b20 on redis/redis-adbc-driver main.
The candidate's Redis Software cluster and helper checks passed. The prefetch qualification record records full race-suite, lifecycle, output-ownership and on/off validation. CUDA validation and its limits are recorded separately.
Tagged-source tests passed. Release validation and upload succeeded: all 20 validation jobs passed, covering the three native builds, ten packaged Linux distribution checks, three installed-wheel client/TLS checks, three DuckDB extension platforms and distribution checks. All 33 release assets are attached. No Go-test skip override or release rerun was used.
Automatic PyPI Trusted Publishing succeeded. All three PyPI 0.0.21 wheels match the verified GitHub release wheel hashes. All 33 downloaded assets match their GitHub API SHA-256 digests, and all 16 checksum files verified. Wheel metadata reports version 0.0.21 and includes native libraries. The downloaded macOS library loaded and exported AdbcDriverInit; this final offline check did not connect to Redis. Live Redis validation is covered by the tagged workflows.
Install from PyPI:
python -m pip install 'adbc-driver-redis[dbapi]==0.0.21'
python -m adbc_driver_redis doctor --offlinePrefetch · GPU scope and requirements · Installation · Full changelog
v0.0.20 — SQL, Arrow and integration distribution updates
Redis ADBC v0.0.20 expands SQL and Arrow support and distributes the native driver, Python wheels, CLI, managed installer, dbc packages and DuckDB extension through the validated release pipeline.
- Add generated keys, UUIDs, data-modifying CTEs, indexes and materialized views.
- Support keyed inserts and opt-in TRUNCATE for adopted tables.
- Extend Arrow normalization and result layouts, Decimal256 validation, key-source streaming and bounded decoded-input handling.
- Improve cursor cleanup, ADBC metadata and partition validation, and SQLAlchemy/DBI compatibility.
- Add authenticated TLS integration coverage, expanded integration acceptance suites and installation guides for 50 tools.
- Publish Python wheels automatically after native, platform, DuckDB and wheel-client release validation succeeds.
Python wheels require Python 3.12+. Native packages support Linux amd64/arm64 with glibc 2.28+ and macOS arm64 on macOS 12+. Redis Search/Query Engine is required; transactions and Redis Flex remain unsupported. Substrait work in this release is design documentation, not implemented support. See each integration guide for its tested versions and scope.
Install Python after the release workflows complete:
python -m pip install 'adbc-driver-redis[dbapi]==0.0.20'
python -m adbc_driver_redis doctor --offlineSource: ed29df91144bded0a27a825092bb477ebf830a4a. Tagged release validation passed all 20 validation jobs and uploaded 33 assets. The initial amd64 toolchain dependency install stalled; the maintainer approved restarting unfinished jobs on a fresh runner. The retry used the same tagged source and standard test settings, with no Go-test skip override. Automatic PyPI publication succeeded. All asset digests and checksum files verified; all three PyPI wheels match their GitHub release hashes. A fresh macOS arm64 installation on Python 3.15 loaded the native driver and exported AdbcDriverInit; this final offline check did not connect to Redis or test manifest discovery. Live Redis validation is covered by the tagged workflow.
Setup · Native installation · Integration installation · Full changelog
v0.0.19 — native installation and connection diagnostics
Redis ADBC v0.0.19 makes installing and configuring the driver easier and expands the documented Python integration paths.
- Add a native installer with automatic platform detection, SHA-256 verification, stable library paths, manifest registration, upgrade and uninstall. Existing manifests require explicit replacement and are restored on uninstall; unrelated or modified files are preserved.
- Add
python -m adbc_driver_redis doctorfor native loading, connection and manifest diagnostics, including--ca-filefor verified private-CA TLS. - Add flat Linux amd64/arm64 and macOS arm64 packages and a generated catalog for dbc onboarding. Public registry publication and Columnar signing still require the Foundry maintainers;
dbc install redisis not yet advertised as available. - Add dedicated integration guides and expanded pandas, Polars, Dagster, Airflow, Prefect, dlt and Ray examples/coverage. See the tool-specific guides for version requirements and test scope.
Python wheels require Python 3.12+; Python 3.14 is the current integration target, with a separate compatible Superset environment. Native releases support Linux amd64/arm64 with glibc 2.28+ and macOS arm64 on macOS 12+. Redis Search/Query Engine is required. Transactions and Redis Flex remain unsupported.
Install Python:
python -m pip install 'adbc-driver-redis[dbapi]==0.0.19'
python -m adbc_driver_redis doctor --offlineInstall the native driver from the checksummed installer asset attached to this release:
python3 install_driver.py install --version v0.0.19See setup, native install/upgrade/uninstall, DuckDB 2.0 setup, and dbc distribution status.
Source: 9f18518e1daddd209c23cb58cfadaef07b66fb39. PR #242's build, topology/architecture, distribution and wheel checks passed. Tagged native, distribution and wheel-client checks passed. All three wheels are published on PyPI. A fresh Python 3.14 installation loaded native v0.0.19 and passed live doctor and both manifest discovery checks.
v0.0.18
Highlights
- Add native Arrow IPC import/export through ADBC and stream eligible
UNION ALLqueries incrementally (#458, #462). - Expand SQL query forms, DROP lists, stored arrays and set-returning functions; improve wide Arrow projections and preserve exact range-scan plans (#453, #455).
- Add ADBC row-count statistics, bounded authoritative HASH counting, and retained result slicing and producer-owned temporary sources (#488, #492). Statistics remain marked approximate because concurrent changes are not isolated.
- Require RESP3 negotiation and reject RESP2 connections (#489). Clients configured for RESP2 must migrate before upgrading.
- Improve in-flight cancellation, transport deadlines, partition/cursor cleanup, recovery lifecycle and execution diagnostics (#492, #495, #496).
- Make Redis Software with verified active 4x Query Performance Factor the default development environment, and harden native TLS/Search readiness checks (#456, #494).
Installation and compatibility
Native packages target Linux amd64/arm64 with glibc 2.28+ and macOS arm64 12+. The release workflow attaches native libraries, CLI tarballs, Python wheels and SHA-256 checksum files after its checks pass. Go source builds require Go 1.26.10 or newer. The base Python package requires Python >=3.12 and ADBC driver manager >=1.12.0; optional integrations have additional requirements.
Redis with Search and RESP3 is required. Writes remain autocommit-only. Transactions, rollback and atomic source snapshots are unavailable. Retained mode freezes completed query output and can observe mixed-time input during capture. Driver-level transaction and consistency work was removed from scope (#500). The Active-Active migration work in #497 is a design proposal, not shipped automatic migration functionality.
Known limitation: intermittent EOFs remain under investigation for a local macOS/Docker three-shard Redis Software environment (#498, #502). The client-only reproduction also observes EOFs, but the root cause is not established. No unsafe write retries or broadened test skips were added to hide these failures.
See driver reference, ADBC statistics, and retained results.
Source and validation
Source: 8cf1efaa03e0275a7bcc0fbfef6156b5944e0bfc (merged #502 on main).
Candidate validation:
- Full Test matrix: all ten jobs passed, including all eight Redis/architecture combinations. Each ADBC suite reported 383 passed, 7 skipped and 3 subtests passed. All four Software fixtures verified active 4x QPF before tests. Expected capability skips are retained in the workflow logs; passing does not imply those capabilities are supported.
- DuckDB extension platform checks: Linux amd64, Linux arm64 and macOS arm64 all passed.
Release binaries passed all 17 jobs: three native builds, ten packaged Linux distribution checks, three installed-wheel client/TLS checks and final upload. All 18 assets are attached (three wheels, six tarballs and their SHA-256 files). Downloaded assets matched all nine checksums; wheel metadata reports 0.0.18 and includes the native libraries. Wheel checksum files retain their build-time dist/ prefix; resolve the named file by basename when verifying a flat download directory. No additional Go test skip pattern was requested. PyPI publication completed successfully. All three 0.0.18 platform wheels are available, and their PyPI SHA-256 digests match the verified GitHub release wheels.
Full changelog: https://github.com/alberttwong/redis-adbc-driver/compare/v0.0.17...v0.0.18
Migrated from the original v0.0.18 release, originally published 2026-10-22T23:44:25Z. Original binary assets are unchanged. The destination tag points to a new root commit with sanitized source and unchanged driver code; earlier Git history was omitted.
Public source snapshots omit private operational artifacts and redact deployment identifiers. Driver code and the original release binaries remain unchanged. See PUBLIC_SANITIZATION.md.
v0.0.17
Highlights
- Add portable ADBC
ExecutePartitions/ReadPartitionfor independent workers, with typed parameter bindings, bounded live row-range scans, selective range pruning and optional balancing (#365, #373, #385, #388). - Add opt-in retained partition results: bounded global SQL execution, immutable Arrow output and independently retryable replay, with tenant namespaces, quotas, expiry, revocation and reader-only credentials (#423, #424, #430).
- Improve ordinary and partition scan throughput and memory use through incremental Arrow construction, faster Search reply decoding, batched HMGET, direct wide-HASH projection decoding and integer aggregate filters. Add an optional accounted-byte budget for streamed batches (#404, #412, #416, #419, #422, #426, #428).
- Harden Search backfill/schema convergence, ALTER cleanup, TLS trust roots and Cluster redirects. Expand installed-wheel coverage across Redis Software, OSS Cluster, native macOS and independent TLS workers (#381, #389, #391, #394, #399, #402, #414, #418).
- Fix DuckDB 2.0 scanner integration, extension TLS loading and scanner ABI version handling (#382, #386).
- Refresh dependencies, require Go 1.26.10 for source builds, and align the Python 3.14 validation environment and manifest-derived wheel build requirements (#438–#441). Superset keeps its own SQLAlchemy/PyArrow compatibility constraints.
Installation and compatibility
Native packages target Linux amd64/arm64 with glibc 2.28+ and macOS arm64 12+. The release workflow attaches driver libraries, CLI tarballs, Python wheels and SHA-256 files after its checks pass. DuckDB extension binaries follow their separate workflow.
The base Python package declares Python >=3.12 and ADBC driver manager >=1.12.0. Optional integrations have their own dependency requirements; the validation harness uses Python 3.14. Install Superset in its own environment because its SQLAlchemy/PyArrow constraints differ from the other extras.
Redis with Search is required. Writes remain autocommit-only; rollback is unsupported. Live partitions remain the default and do not share a snapshot. Retained mode freezes completed query output, not an atomic Redis source snapshot; concurrent source changes can produce mixed-time input. Retained mode requires explicit namespace, persistence, ACL and resource-limit configuration. See live partitions, retained results and Python integrations.
Source and validation
Source: 9cc8e25c0f0149b19d2d76446fa45a280317c8d4 (merged #441 on main).
The release workflow, attempt 2, passed all three native platform jobs, all ten packaged Linux distribution checks, all three installed-wheel client/TLS checks, and final asset upload. All 18 versioned assets (three wheels, six tarballs and their SHA-256 files) are attached below.
Attempt 1 failed in TestHashScan on Linux amd64: the test assumed an unordered three-HASH sample excluded the fixture's nonnumeric age. Attempt 2 reran the failed jobs against the same original tag with all tests enabled and no additional skip pattern. The tag and source commit were not changed. PR #444 makes that test deterministic for future runs; its focused integration test passed 25 consecutive local runs.
PyPI publication is a separate workflow and is not yet claimed for this version.
Full changelog: https://github.com/alberttwong/redis-adbc-driver/compare/v0.0.16...v0.0.17
Migrated from the original release, originally published 2026-10-22T16:23:50Z. No uploaded binary assets were copied. This tag contains a sanitized standalone source snapshot of original commit 9cc8e25c0f0149b19d2d76446fa45a280317c8d4; original commit history was omitted. Private operational artifacts and deployment identifiers were removed; driver code is unchanged. See PUBLIC_SANITIZATION.md.
v0.0.16
Highlights
- Add direct partitioned Dask DataFrame reads through ADBC, with explicit query parameters and Arrow schemas, worker-local connections, and preservation of nullable types and empty partitions (#357).
- Add the
daskPython extra, a runnable two-process Distributed example, and installation, partition sizing, and consistency guidance. - Add the standalone
redis_adbcDuckDB extension and make DuckDB 2.0 the primary build/runtime target (#344, #354). Its extension build and distribution follow the separate DuckDB workflow; the native driver release pipeline does not package DuckDB extension binaries. - Improve release cluster readiness checks, bound distribution compatibility validation, reduce repeated CI builds, and add PyPI Trusted Publishing and package descriptions (#339, #345–#350).
Installation and limitations
Install the published wheel and Dask dependencies:
python -m pip install 'adbc-driver-redis[dask]==0.0.16'The Dask helper requires caller-defined disjoint partitions and an explicit Arrow schema. Each partition is materialized in worker memory; reads and retries have no shared snapshot. Install matching packages and the native driver on every worker. See the integration guide.
Redis with Search is required. Writes remain autocommit-only; transaction rollback is unsupported. Platform wheels target Linux x86-64/ARM64 with glibc 2.28+ and macOS ARM64 12+. Native libraries, CLI tarballs, wheels, and SHA-256 files are attached below.
Validation and publication
Source: c737cb8fff88855bc1c12e0d48f3b4dd3c46a364 (merged #357 on main).
Local Dask validation passed 24 Python integration tests, with 2 explicitly optional tests skipped (DuckDB extension download and isolated Superset environment). The Distributed example passed with partition sizes [1, 2, 2, 0, 1]. Minimum-version validation passed all 8 Dask tests using Dask/Distributed 2024.5.0 and PyArrow 14.0.2. See the verification record.
The release workflow passed: Linux amd64/arm64 builds and Redis Software/OSS cluster tests, macOS arm64 build/load/unit checks, all ten packaged Linux distribution checks, and final asset upload succeeded.
PyPI Trusted Publishing passed. Version 0.0.16 exposes the Dask extra, and all three PyPI wheel hashes match the validated release artifacts. A clean installation of the macOS release wheel also passed the two-process Dask query example against disposable Redis using its bundled native library.
Full changelog: https://github.com/alberttwong/redis-adbc-driver/compare/v0.0.15...v0.0.16
Migrated from the original release, originally published 2026-10-20T22:04:05Z. No uploaded binary assets were copied. This tag contains a sanitized standalone source snapshot of original commit c737cb8fff88855bc1c12e0d48f3b4dd3c46a364; original commit history was omitted. Private operational artifacts and deployment identifiers were removed; driver code is unchanged. See PUBLIC_SANITIZATION.md.
v0.0.15
Highlights
- Add platform Python wheels with
adbc_driver_redisDB-API helpers, theadbc_driver_redissTLS alias, Polars URI loading, option enums, and Redis/rediss manifest installers (#338). - Add SQLAlchemy 1.4/2.x dialects and Superset engine specs with catalog reflection and PostgreSQL time grains. Add a Dagster IO manager for Arrow, pandas, polars, staged table replacement, and partition retries (#338).
- Document and test DuckDB's current ADBC scanner integration, including attached-table autocommit writes, bulk ingestion, schemas, read-only attachments, and rejection of explicit transactional writes (#338).
- Expand PostgreSQL SQL, session, metadata, catalog, and error compatibility, including byte-order COLLATE support and application HASH discovery (#231, #234, #332, #333).
- Add bounded multicore scans, shared worker budgets and measured worker tuning; improve exact aggregate spill merging and ordered top-k execution (#223, #225, #233, #304, #326, #328–#330).
- Improve Redis Software setup/recovery checks, benchmark validation/cleanup, and client integration documentation.
Upgrade notes and limitations
- Unquoted SQL identifiers fold to ASCII lowercase; quote legacy mixed-case object names. Small integer literals infer INTEGER before BIGINT/NUMERIC. Default descending sorts place NULLs first.
- Text read as TIMESTAMP without time zone now keeps the written wall-clock time and ignores a validated offset or zone, as PostgreSQL does. Previously stored timestamps are not rewritten. Review stored view/default/check expressions when upgrading from v0.0.14.
- Strings use PostgreSQL C/POSIX byte ordering; locale-aware collations remain unsupported.
- Writes remain autocommit-only. SQLAlchemy rollback cannot undo writes. Dagster table replacement has a DROP/RENAME window, and partition DELETE/append can partially apply. Real multi-statement transactions are not implemented.
- Python wheels are distributed here; PyPI publication and driver-registry registration are not configured. macOS release wheels target macOS 12+ with Go 1.26. Native manifest paths are absolute; rerun installation if the library moves.
- Redis Search is required. Redis Flex and unsupported Redis-compatible products remain outside the supported scope.
Validation and downloads
Release source: 11faf538d0499301c1cb79c7a9ab3e195221a9ea on main (merged #338).
Local #338 validation on macOS arm64 with a disposable Redis 8.6.2 instance passed 17 Python integration tests, 3 separate Superset/SQLAlchemy 1.4 checks, 4 clean installed-wheel tests, the Go unit suite, and focused native collation/catalog regressions. The main Python run skipped Superset because it was tested separately. Browser-rendered charts, a TLS handshake, and Columnar's separate DuckDB extension were not exercised.
PR CI was still running at publication. Publishing starts the release workflow, which builds Linux amd64/arm64 and macOS arm64 native tarballs, CLI tarballs, Python wheels, and SHA-256 files. Downloads appear after its build/test/distribution validation jobs succeed; publication itself does not mean those checks passed.
See the verification record and release workflow for evidence and build status.
What's Changed
- docs: clarify unsupported Redis-compatible products by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/222
- Use multiple cores for individual aggregate queries by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/223
- Share an adaptive worker budget across aggregate queries by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/225
- Refresh README performance measurements on the current driver by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/228
- Default to Redis Software and correct QPF benchmarks by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/229
- Replace README latency table with measured driver worker scaling by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/230
- Read text as a number, boolean, date or time as Postgres does, with its errors by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/231
- Add measured aggregate worker autotuning with bounded exploration by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/233
- Ignore an offset or zone in text read as a TIMESTAMP, as Postgres does by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/234
- Create the Redis Software cluster once its API answers by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/235
- Fix 15 PostgreSQL compatibility, ADBC, and pagination bug reports by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/281
- Add starter Postgres differential harness (#236) by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/282
- Test four Redis Search environments in parallel by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/284
- Expand differential corpus and fixture coverage (#283) by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/287
- Fix concurrent ALTER failures on sharded Redis Software by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/288
- Stabilize worker tuning and document stopped performance validation by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/289
- Preserve legacy mixed-case column resolution by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/290
- docs: list data tool integrations and testing status by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/291
- Isolate ClickBench trials with hard memory limits by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/292
- Bound exact ClickBench aggregates and recover from Redis group limits by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/304
- Fix ClickBench follow-up query plans, FinOps preflight, and typed references by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/314
- Reduce Q19 count-group overhead and preserve timeout stage evidence by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/315
- Fix ClickBench cleanup order for runner security groups by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/324
- Validate ClickBench LIMIT ties and bound cleanup polling by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/325
- Improve ClickBench exact execution and retain live diagnostics by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/326
- Parallelize exact top-K fallback for ClickBench Q26 by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/328
- fix: parallelize residual ordered top-k by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/329
- perf: avoid redundant aggregate spill merge passes by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/330
- fix: verify terminated runners without VPC metadata by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/331
- Add PostgreSQL session, metadata, and query compatibility by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/332
- Complete adopted text encoding, foreign indexes, and PostgreSQL catalogs by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/333
- docs: compare ADBC and Redis connectivity across 20 data tools by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/334
- docs: remove superseded README integration table by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/335
- docs: show Redis connection solution paths with icons by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/336
- docs: add project-grounded AI-native SDLC records by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/337
- Add Python packaging and analytics client integrations by @alberttwong in https://github.com/alberttwong/redis-adbc-driver/pull/338
Full Changelog: https://github.com/alberttwong/redis-adbc-driver/compare/v0.0.14...v0.0.15
Migrated from the original release, originally published 2026-10-20T17:54:52Z. No uploaded binary assets were copied. This tag contains a sanitized standalone source snapshot of original commit 11faf538d0499301c1cb79c7a9ab3e195221a9ea; original commit history was omitted. Private operational artifacts and deployment identifiers were removed...