v1.1.0

What's Changed

• Swift/iOS binding over the native core: feature parity + docs (#28) by @Davidobot in #47

Full Changelog: v1.0.0...v1.1.0

v1.0.0

What's Changed

• feat(migrate): lodedb migrate toolkit for existing vector stores by @Davidobot in #36
• docs(agents): document the migrate toolkit and guard its source-provider imports by @Davidobot in #37
• feat(late-interaction): multi-vector MaxSim retrieval for visual-document RAG (#25) by @Davidobot in #38
• native-core: add Rust core migration by @Davidobot in #40
• bench: widen native-vs-Python lexical-search ratio gate to 1.10 by @Davidobot in #43
• release: unblock + harden the v1.0.0 publish (swift-tools-version 5.9 + decouple PyPI from xcframework) by @Davidobot in #44

Full Changelog: v0.4.0...v1.0.0

v0.4.0

What's Changed

• feat(integrations): add PrivateGPT vector-store provider by @Davidobot in #21
• feat(embeddings): default to ONNX Runtime with a PyTorch fallback by @Davidobot in #29
• fix(onnx): honest doctor report + default to the CPU provider on Apple Silicon by @Davidobot in #30
• fix(ops): harden local runtime production defaults by @Davidobot in #32
• Multimodal embedding: CLIP image+text preset, custom embedders, named vector spaces by @Davidobot in #26

Full Changelog: v0.3.0...v0.4.0

v0.3.0

What's Changed

• Memory-integration benchmark: LodeDB vs framework default stores by @Davidobot in #20
• feat(engine): add opt-in WAL commit mode by @Davidobot in #23

Full Changelog: v0.2.1...v0.3.0

v0.2.1

What's Changed

• Improve MCP search for RAG; detect Windows GPU torch in doctor by @Davidobot in #15
• feat(integrations): add LlamaIndex VectorStore and PropertyGraphStore adapters by @Davidobot in #7
• docs: add integrations roadmap and progress tracker by @Davidobot in #8
• feat(mcp): one-command MCP install for coding assistants by @Davidobot in #19
• feat(integrations): add mem0 adapters by @Davidobot in #16

Full Changelog: v0.2.0...v0.2.1

v0.2.0

What's Changed

• feat: knowledge-graph and memory backend (vector-in, vector-only index, filter planner, enumeration, graph layer) by @Davidobot in #11
• Add hybrid lexical + vector search (BM25 + RRF) by @Davidobot in #14

Full Changelog: v0.1.2...v0.2.0

v0.1.2

What's Changed

• Single-writer cross-process persistence lock by @Davidobot in #3
• Crash-atomic commits via generation-addressed root manifest by @Davidobot in #5
• bench: local search_many batched-retrieval throughput benchmark by @Davidobot in #9
• Keep incremental commits O(changed): defer SIMD repack + drift, journal raw text by @Davidobot in #6
• First-class opt-in Apple-GPU (Metal/MPS) exact scan by @Davidobot in #4
• Push query filters into the batch scan as a shared allowlist by @Davidobot in #10
• Add metadata filter predicate operators by @Davidobot in #2

New Contributors

• @Davidobot made their first contribution in #3

Full Changelog: v0.1.1...v0.1.2

v0.1.1

What's Changed

• perf: implement O(changed) in-place GPU row patching by @the-alt-dev in #1

New Contributors

• @the-alt-dev made their first contribution in #1

Full Changelog: v0.1.0...v0.1.1

v0.1.0

Full Changelog: https://github.com/Egoist-Machines/LodeDB/commits/v0.1.0