Workflow for adding new datasets

[LinguList]

I will soon make this clearer, but the general information I should give you all here is: we will

1. separate data from code as far as possible, using code to lift data, but not mixing analysis code with code for data curation
2. make separate repositories for datasets, like Baxter1992, Wang1980, where data can be found in the same structure in the same folder, and we model datasets as Python packages, so you can install them and import them in Python scripts (this makes analysis replicable!)

So please already get used to organizing individual datasets in individual repos, and using minimal code to convert data from raw formats to our exchange (cldf-like) format.