reproducible executable environments
Binder is a collection of tools for building and executing version-controlled computational environments that contain code, data, and interactive front ends, like Jupyter notebooks. It's 100% open source. We maintain a small cluster for public use, but it's also easy to deploy the system yourself.
This repository does not contain any actual code, but serves as a reference for Binder information, and a place to post issues or questions about the project. All Binder components are written in node.js but an earlier version written in python can be found on the legacy branch of this repository.
See docs.mybinder.org for official documentation.
At a high level, Binder is designed to make the following workflow as easy as possible
- Users specify a GitHub repository
- Repository contents are used to build Docker images
- Deploy containers on-demand in the browser on a cluster running Kubernetes
Common use cases include:
- sharing scientific work
- sharing journalism
- running tutorials and demos with minimal setup
- teaching courses
Binder is implemented through a collection of NodeJS modules, each of which can be independently tested and versioned. The key components are:
binder-buildbuild Docker images from repository contentsbinder-deploy-kubernetesdeploy images on a Kubernetes clusterbinder-controlCLI for setting up binder components for a deploymentbinder-clientCLI and library for interacting with a binder deploymentbinder-webweb frontend for a Binder deployment
We maintain a public Binder cluster at mybinder.org running on Google Compute Engine, supported by HHMI Janelia Research Center, and designed for open source and open science projects. You just need to specify a GitHub repository, and you'll get a badge to embed in your project README that launches the environment. Head to mybinder.org to try it out.
We've also made it easy to setup a custom Binder deployment on your own compute infrastructure. This is a great idea if you need guarenteed availability (e.g. for a course), want to use an existing compute cluster, or need access to private data. It's also a great way to understand the system and start contributing new features! See the devs section of docs.mybinder.org to get started.
We welcome community contributions! You can submit issues or pull requests to the repository for the component you're interested in working on, e.g. if you have an idea for improving how dependencies are resolved, open an issue on binder-build. But if you're unsure, you can just open an issue on this repository.