Hey Rustaceans!

If you’ve ever built geospatial applications, you know that GDAL is the undisputed king. But you probably also know the pain of wrestling with gdal-sys, dealing with complex C/C++ toolchains, and ending up with bloated Docker images just to read a GeoTIFF.

Today, we are super excited to announce the v0.1.0 release of OxiGDAL — a production-grade Geospatial Data Abstraction Library written entirely in 100% Pure Rust.

No C/C++ dependencies. No Fortran. Just a clean cargo add oxigdal and you're good to go.

Why we built it & What it does:
We wanted a modern, memory-safe geospatial engine designed for the cloud-native era. OxiGDAL spans over 68 workspace crates and includes:

• Zero-Dependency Build: Compiles to a statically linked binary. Our production Docker images are under 50MB.
• Blazing Fast SIMD: We heavily utilize AVX2/NEON for coordinate transformations and raster math. (e.g., transforming 1M points takes < 10ms).
• Arrow-Native Vector Data: Reads GeoParquet natively. It's built on top of the Rust Arrow ecosystem, bypassing traditional serialization overhead.
• WASM First-Class Support: Because it's pure Rust, you can parse Cloud Optimized GeoTIFFs (COGs) and run heavy geospatial algorithms directly in the browser.
• Enterprise / Cloud-Native: Built-in pure Rust crypto (rustls, ring), Raft-based clustering, and native streaming integration (Kafka, Pub/Sub).

Quick example (CRS Transformation):

use oxigdal_proj::{Crs, Transformer};

// No C-bindings involved!
let wgs84  = Crs::from_epsg(4326)?;
let utm54n = Crs::from_epsg(32654)?;   
let tf     = Transformer::new(&wgs84, &utm54n)?;

// SIMD-vectorized batch transform
let (easting, northing) = tf.transform(139.7671, 35.6812)?;

We just published a Medium article detailing the architecture and motivation behind the project:
Introducing OxiGDAL: The Pure Rust, Cloud-Native Successor to GDAL

You can check out the source code and documentation here:
GitHub: https://github.com/cool-japan/oxigdal
Crates IO: cargo add oxigdal

We’d love to hear your feedback, issues, or PRs. Let's make the future of GIS memory-safe!

2 posts - 2 participants

Read full topic