GeoLibre 1.0 Release Notes

GeoLibre 1.0 is a stable prototype of a cloud-native Geographic Information System (GIS) platform designed for visualizing, exploring, and analyzing geospatial data. By leveraging a modern web stack, it provides a responsive workspace that operates across desktop and web environments, reducing the friction associated with traditional heavy GIS software.

Core Architecture and Tech Stack

GeoLibre is built on a high-performance, cloud-native stack to ensure fast local and remote data processing. The primary technologies include:

  • Frontend Frameworks: React and TypeScript.
  • Map Rendering: MapLibre GL JS and deck.gl.
  • Data Engine: DuckDB-WASM Spatial for in-browser SQL analysis.
  • Application Shell: Tauri, which enables the same workspace to run as both a desktop application and a web app.

Key Features and Capabilities

Geospatial Data Integration

GeoLibre supports a wide array of local and remote data formats, allowing users to load vector and raster data, inspect attributes, and apply data-driven symbology. Supported formats include:

  • Cloud-Native Formats: GeoParquet, FlatGeobuf, PMTiles, and Zarr.
  • Web Services: XYZ, WMS, WFS, WMTS, ArcGIS, and STAC services.
  • ** uma Raster Formats:** COG (Cloud Optimized GeoTIFF), GeoTIFF, and MBTiles.
  • 3D and Specialized Data: LiDAR, Gaussian splats, and 3D Tiles.
  • Databases: DuckDB and PostgreSQL.

Analysis and Processing Tools

GeoLibre provides integrated tools for both vector and raster analysis, as well as a dedicated SQL workspace.

  • SQL Workspace: Users can run DuckDB Spatial SQL directly in the browser. The system automatically wraps bare URLs into the matching reader and streams remote files via HTTP range requests.
  • Vector Processing: Common geometry tools (buffer, centroids, convex hull, dissolve, bounding box, simplify, clip, intersection, difference, and union) are powered by Turf.js in the browser, with an optional GeoPandas sidecar engine for more intensive tasks.
  • Raster Processing: Tools such as hillshade, slope, aspect, reproject, resample, and contour are handled by a rasterio Python sidecar.
  • Geoprocessing: The Whitebox toolbox is available via an optional Python sidecar for batch processing.

Extensibility and Integration

GeoLibre includes a built-in plugin marketplace for installing and updating external plugins. Built-in plugins include support for Overture Maps, LiDAR, GeoAgent, and GeoEditor. Additionally, the platform can be embedded into Jupyter notebooks via the geolibre Python package, providing a leafmap-style API (add_geojson, add_tile_layer, add_cog) that synchronizes UI edits back to Python.

Deployment and Sharing

GeoLibre offers a live browser demo hosted on GitHub Pages. This version is a static site that processes data client-side, ensuring privacy by design. While the browser version supports URL-based layers and DuckDB-WASM Spatial loading, certain operations—such as local MBTiles reads and filesystem save/open operations—require the installed Tauri desktop app.

Projects can be shared via .geolibre.json files. The demo supports several URL parameters for embedding and sharing, including layout=compact for narrow embeds and maponly for a chrome-free, map-only view.

Community Feedback and Technical Observations

User discussions highlight both the potential and the limitations of the current 1.0 prototype:

  • Convenience vs. Power: Users have noted that browser-based GIS is significantly more convenient than traditional tools like QGIS, particularly for working with public datasets hosted on ArcGIS Online.
  • Performance Limitations: Some users reported IO errors in the web version and performance degradation when loading large files (e.g., files >1GB), noting that the platform may need further optimization for large-scale data imports.
  • Comparison to ArcGIS Online: The platform is viewed by some as a viable, subscription-free alternative to the ArcGIS Online Map Viewer, potentially benefiting non-profits.
  • Stack Validation: Developers have noted that the combination of MapLibre, DuckDB, and PMTiles is a highly performant and pleasant stack for modern geospatial work.

Sources