World-wide ranking of geographic locations based on OpenStreetMap tile logs. Updated weekly. Aggregated over the past 52 weeks to smoothen seasonal effects. For any location on the planet, up to ~150m/z18 resolution.
cmd/webserveris the OSMViews webserver.cmd/osmviews-builderis the pipeline that computes the data.docscontains further documentation.
Client libraries are maintained in separate repositories:
- Python: brawer/osmviews-py
- Rust: brawer/osmviews-rs
Contributions are welcome — see CONTRIBUTING.md.
The techniques used by cmd/osmviews-builder — and by the
osmviews-rs and
osmviews-py client libraries — are
written up as a defensive publication. It doubles as a high-level tour of
the whole system: the level-embedding tile key, the streaming per-period sort
and cross-period merge, the constant-memory raster construction with inline
overviews, and the Cloud-Optimized GeoTIFF layout.
Method for Memory-Bounded Construction of a Globally Complete, High-Zoom-Level Cloud-Optimized GeoTIFF from Tile-Access Logs. Technical Disclosure Commons, 2026.
Download the PDF · source and build instructions
It is published to establish prior art and keep these techniques free to use.
Code: MIT — see LICENSE.
The defensive publication under
docs/defensive-publication/ — the LaTeX source
and the rendered PDF — is licensed CC BY 4.0.
The OSMViews raster this pipeline produces is released into the public domain under CC0 1.0.
This repository is REUSE compliant: every file
declares its copyright and license, either in an SPDX header or via
REUSE.toml.