Adopt Community Package for Data Fetching

[mandli]

Found this community supported package for fetching data files instead of using the direct libraries urllib and the such. Provides:

• Way better error handling than we currently have. No more data files existing that acutally contain http fetch request failures rather than data.
• Built-in abilities to handle multiple types of data, e.g. zip.
• Can fetch using more than just URLs, e.g. DOIs.
• Also handles multiple protocols.

Pooch

[rjleveque]

Thanks @mandli! Looks like a great package, easy to install and use and I like that it can download directly from Zenodo, for example.