Kjell Wooding

20_newsgroups is created on a make_test. Should be removed, or confined to CI only

right now it doesn't. Found out the hard way when implementing extra_base

I should be able to download a LICENSE or README from a URL and add them to a datasource

Listing all the raw files in a datasource would greatly help the Makefile do the right thing if one of them changes (or is missing)

when calling add_file, check the filename/hash of the file against the current file list before adding. (via mark)

and make data should go data/processed -> data/processed

Goal: ship a tarball that excludes the git repo info (and checks that it's not included)

test_notebook_generic_edge is super slow. Change it to download something smaller