Feature request: dask book

This came up in the chat of the dask tutorial at SciPy 2020.

I think it's a great idea. Mostly because I love reading books on planes. There is something nice about slowly reading through code on paper to understand how the API works etc.

The FastAI book could offer a good template https://github.com/fastai/fastbook and is a good supplement to courses/videos/tutorial/blog etc.

I believe there are a ton of spark books e.g. https://www.oreilly.com/library/view/learning-spark-2nd/9781492050032/. A dask book could take inspiration from these.

The issue proposes a Dask book but names no repository files, tests, or implementation entry points. Start by reading the linked FastAI book and Learning Spark reference to understand the proposed format and scope. Done would require an agreed book outline and a concrete plan for producing it.

Written by the indexing model from the issue text.