1 unstable release

0.0.3 Feb 18, 2023
0.0.2 Apr 26, 2022
0.0.1 Apr 19, 2022

#20 in #web-gpu

Download history 13/week @ 2024-02-19 28/week @ 2024-02-26 3/week @ 2024-03-04 10/week @ 2024-03-11 3/week @ 2024-03-18 8/week @ 2024-03-25 67/week @ 2024-04-01

89 downloads per month
Used in 2 crates

MIT/Apache

475KB
12K SLoC

Native Maps for Web, Mobile and Linux
A map rendering library written in Rust.

Project State

This project is in a proof-of-concept state. The proof of concept is done except for text rendering. The Rust ecosystem is suited very well for this project.

In the future, this project could be adopted and supported by Maplibre to implement a next-gen mapping solution.

📰 We recently released a paper about maplibre-rs called maplibre-rs: toward portable map renderers!

Description

maplibre-rs is a portable and performant vector maps renderer. We aim to support web, mobile and desktop applications. This is achieved by the novel WebGPU specification. Plenty of native implementations are already implementing this specification. On the web, it is implemented by Firefox, Chrome and Safari. There are also standalone implementations that directly use Vulkan, OpenGL or Metal as a backend. Those backends allow maplibre-rs to run on mobile and desktop applications.

Rust is used as a Lingua-franka on all platforms. This is made possible by WebAssembly, which allows us to use Rust for web development.

The goal of maplibre-rs is to render maps to visualize data. Right now the goal of maplibre-rs is not to replace existing vector map renderers like Google Maps, Apple Maps or MapLibre. The current implementation serves as a proof-of-concept of the used technology stack. It is unclear whether the high-performance requirements of rendering maps using vector graphics are achievable using the current stack.

Talk: World in Vectors

(External Link)

(Older Talk on YouTube)

Demos

Current Features

  • Runs on Linux, Android, iOS, macOS, Firefox and Chrome
  • Render a vector tile dataset
  • Simple navigation powered by winit
  • Multithreaded on all platforms
  • Querying feature data

Missing Features

  • Rendering Text
  • Per-Feature Rendering
  • Rendering:
    • Labels
    • Symbols
    • Raster data
    • 3D terrain
    • Hill-shade (DEM)
  • Collision detection
  • Support for:
    • GeoJSON
  • API for:
    • TypeScript
    • Swift
    • Java/Kotlin

Building & Running

Clone the project

git clone https://github.com/maplibre/maplibre-rs.git

Build and run it on a desktop

cargo run -p maplibre-demo

More information about running the demos on different platforms can be found here.

Rust Setup

Install rustup because this is the recommended way of setting up Rust toolchains.

The toolchain will be automatically downloaded when building this project. See ./rust-toolchain.toml for more details about the toolchain.

API Documentation

This generates the documentation for this crate and opens the browser. This also includes the documentation of every dependency.

cargo doc --open

You can also view the up-to-date documentation here.

Book

The maplibre-rs book features a high-level overview over the project from a user and development perspective.

RFCs

We established an RFC process which must be used to describe major changes to maplibre-rs. Current RFCs can be browsed in the book.

Citing

If you wish to cite this project in a scientific publication use the following format:

@article{maplibre_rs,
	title        = {maplibre-rs: toward portable map renderers},
	author       = {Ammann, M. and Drabble, A. and Ingensand, J. and Chapuis, B.},
	year         = 2022,
	journal      = {The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences},
	volume       = {XLVIII-4/W1-2022},
	pages        = {35--42},
	doi          = {10.5194/isprs-archives-XLVIII-4-W1-2022-35-2022},
	url          = {https://www.int-arch-photogramm-remote-sens-spatial-inf-sci.net/XLVIII-4-W1-2022/35/2022/}
}

Acknowledgements

The renderer of maplibre-rs is heavily based on the renderer of bevy. Bevy's renderer was forked into this project in order to have a solid and generic base.

Dependencies

~18–61MB
~1M SLoC