Export to Binder
The Binder project has led to the development of reproducible Jupyter notebooks that can be executed, changed, and examined for free on the web. A GitHub repository can be used to specify the environment used, including what Python modules are needed for the Jupyter kernel and what extensions should be installed for JupyterLab before the notebook opens in the browser. A Docker container is generated on the fly in response to the specification in the repository, and once ready the repository is mounted inside the container and the notebook opened.
This is clearly a powerful tool that is ready for use by the wider community, and offers a path to providing full access to the data and notebooks for anyone without the need to create accounts. The Python module that had been developed to extend the Jupyter kernel was extended to support execution without a data server, and packaged using pip so that it could be installed via the mechanism offered by the Binder project. A blog post was written in December, 2018 describing the approach \cite{blog}, and showing the Binder repository with example links to data for calculations \cite{openchemistryjupyter-examples}.