This builds a container that can be used as a base for any notebook deployed to JupyterHub. This container sets up JupyterHub and installs an environment that looks an awful lot like our primary bastion host including operating system version and installed software. It also installs a basic Python 3 version with some standard libraries and JupyterLab extensions.
If you update the version of JupyterHub in this container you need to make sure that it matches the version of JupyterHub installed in the main JupyterHub container.
You could actually run this base notebook directly if you wanted. However, you should really inherit from this and add additional things. The best example of that is the datascience-notebook which is the standard notebook for the UW CIP's JupyterHub.
If you update this container you must tag it and push the changes and the tag to GitHub to force a new build to happen. When the build finishes successfully then you will need to update any container that references this container to reference the new version.