12 posts
OxiGeo 0.2.4 fixes three real user-filed issues — GeoPackage overflow-page cells (#17), GDAL's fractional VRT SrcRect/DstRect values (#18), and VRT mosaic nodata-skip compositing (#19) — plus a new buffered sync-over-fetch driver that makes the browser COG viewer work end to end for the first time, and two large embedded EPSG registry correctness passes.
OxiGeo 0.2.3 fixes GitHub issue #15 — every gdalwarp -of VRT product was rejected at parse time — with a real backward warp engine in oxigeo-vrt, and issue #16 — Dataset::open on a GeoPackage always reported 0 layers — with a new Dataset::layers()/Layer::features() vector API for GeoPackage, Shapefile, and GeoJSON.
OxiGeo 0.2.2 fixes GitHub issue #14 — Dataset::read_band silently returned the whole multi-band image instead of one band, a defect pattern found and fixed in a dozen crates. Also ships 1.45-1.79x faster DEFLATE decoding and zero-allocation typed raster readers.
OxiGeo 0.2.1 is a production-hardening release: a multi-agent defect sweep found 342 bugs and fixed 314, including two CRITICAL silent-corruption decoders (JPEG2000, GRIB2) and a WFS-T fail-open security hole. A new axum gateway serving layer ships, and retiring oxigeo-kafka makes the workspace fully C-toolchain-free.
OxiGDAL 0.1.7 is a production-hardening release: 233 verified defects fixed across 69 crates — a GeoTIFF corruption bug, JPEG2000 spec conformance, real FlatGeobuf/HDF5 codecs, an RBAC bypass — plus a blake3/Ed25519 attestation module and three new WASM demos. 76 crates, 16,909 tests. The sovereign geospatial layer for COOLJAPAN.
OxiGDAL is now OxiGeo. Version 0.2.0 is a rename-only release — functionally identical to v0.1.7 — that republishes all 74 crates, the CLI, and the Python/npm/WASM/mobile bindings under the oxigeo name. Old GitHub URLs redirect; v0.1.7 remains the final OxiGDAL release. The sovereign geospatial layer for the COOLJAPAN ecosystem, now under its permanent name.
OxiGDAL 0.1.6 eliminates rusqlite/libsqlite3-sys and ring from the workspace — SQLite now runs on oxisql-sqlite-compat (Limbo) and TLS on OxiTLS, making the default build 100% C/FFI-free. Plus GPX/KML/TopoJSON formats, non-UTF-8 DBF encoding, Delaunay triangulation, HDF5 v2/v3 superblocks, terrain GLCM/geomorphons/cost-distance, batch QC validators, and W-TinyLFU caching. 78 crates, 18 drivers, 14,605 tests.
OxiGDAL 0.1.5 is a focused fix release: a stray padding field in the RayMarchUniforms WGSL layout was shifting every field by 4 bytes, making the GPU compute kernel read a billion-step max and hang indefinitely on macOS Metal. With the layout corrected, the ray-march GPU/CPU parity test passes in 0.127s. 78 Pure Rust workspace crates, 14,605 passing tests.
OxiGDAL 0.1.4 lands Weiler-Atherton polygon clipping, Karney geodesic area, DE-9IM topology predicates, and marching-squares contours, plus AVX2/NEON SIMD resampling, a completed COPC point-cloud reader, R-tree STR bulk loading + kNN, and a migration from ort to oxionnx for fully Pure Rust ONNX inference. 76 crates, 12,064 passing tests.
OxiGDAL 0.1.2 adds a PMTiles v3 writer with Hilbert tile IDs and content dedup, a full geometry validation/operations module (centroid, area, point-in-polygon, simplify, convex hull), and a unified umbrella crate with 12-format conversion planning. Now 76 Pure Rust workspace crates, 15 format drivers, 10,935 passing tests — zero C dependencies in default features.
Production-grade pure Rust geospatial data abstraction library. 11 format drivers (GeoTIFF/COG, GeoParquet, Zarr, NetCDF, etc.), 211+ EPSG definitions, cloud-native async I/O (S3/GCS/Azure), GPU acceleration (wgpu), WASM + Python bindings. ~480k SLoC, 69 crates, 7,486 passing tests. 10× faster than GeoPandas, <50 MB Docker, no C deps. The sovereign geospatial layer for SciRS2 and the entire COOLJAPAN ecosystem.
The first public release of OxiGDAL: a 100% Pure Rust reimplementation of GDAL with zero C/C++/Fortran in default features. 11 format drivers (GeoTIFF/COG, GeoParquet, Zarr, NetCDF, HDF5, GRIB, JPEG2000, ...), a Pure Rust PROJ with 211+ EPSG definitions, SIMD raster/vector algorithms, cloud-native async I/O, and WASM + Python bindings — across 68 workspace crates.