packages tutorial-- to do list

[jeremymanning]

• need better system for starting packages-- maybe a trello board?
• proposal:
  • let's start with a tello board to outline the key important components
  • then we'll see what we can/should automate

[jeremymanning]

• create a standard requirements file
• create a standard license
• create forks of a template that have application-specific requirements/formats (e.g. psychopy experiment, javascript experiment, new python 2/3 toolbox, jupyter notebooks repo, etc.)
• add/modify docker to support a superset of everything

[jeremymanning]

Documentation:

• What are the key things that every lab project should include?
• Can we define a specific set of instructions or a "lab standard pipeline"
• We should make readthedocs pages for each project

[jeremymanning]

Cookie cutter has an interesting solution: https://github.com/audreyr/cookiecutter-pypackage

[jeremymanning]

It'd be helpful to have an "anatomy of a Python package" document-- e.g. what is the minimum viable set of components needed to consider a collection of files a complete python package?

Then we could add to that:

• What additional requirements (e.g. documentation, testing, license, convenience functions, examples, tutorials, etc.) do we want to always have for lab packages?

[jeremymanning]

Let's come back to this topic after round 1 of our tutorials