This is a first pass at the tech and tool bootstrapping around a reproducible experimental environment. This is mostly focused on science hackers and experimentalists since formal application development and wet-digital hybrid experimentation has its own best practices.

• Create Github Account
• Fork deardooley/agave-mulligan

    - .travis.ci

    - .dockerignore
    + assets
    + data
    - docker-compose.yml
    - discovery.yml
    + docs
    - index.ipynb
    - publish.sh
    + project
    - Dockerfile
    - AUTHORS.md
    - README.md
    - LICENSE

• Set up automatic build on docker hub
• Add static assets like images, etc to assets folder
• Update AUTHORS.md with your project members' names
• Update LICENSE if apprpriate
• Add code to project directory * Source * Tests * Build file
• Open Jupyter and update intro with your project info

• Work out of Jupyter, or on your code.
• Document process in notebook
• Save notebook and commit along with code.
• Branch for new ideas, directions, etc
• Set up .travis.ci to run tests for free
• Add default hello world test data to data dir

• Gitter for conversations and communication logs
• Comment on releases and reference commits.

• list any relevant external sources of data, apps, code, etc in the discovery.yml file
• Update docs to produce valid usage information for your app if appropriate
• Create a release of the project
• Update the resulting Docker build to match the release tag.

• Run publish.sh to register your container as an app for others to use
• Run benchmark.sh to benchmark your app using the sample data in the data folder and/or discovery.yml file.
• Add a "Try it yourself" button to your project readme.