Airtable integration

[rjbergerud]

Not sure how useful this will be. Potential downsides being that we are missing entries in some fields in the airtable (airtable doesn't offer option for making fields mandatory), if the code makes use of them it with throw an error unless we put a bunch of checks everywhere. So far, really just using the 'Name', 'Link', and 'Type' fields.

Benefit is that we can keep a single-source of datasets that we're tracking (in airtable), be able to view this from the airtable notebook, and maybe more quickly dive into one of them if need be.

Potential to add a field to airtable in the future which flags how similar datasets should be processed. But again, not sure if this is the best route, since it's easy to make changes (delete a tag) in airtable that mess things up downstream, and so perhaps those sorts of tags should be kept in git.

[rjbergerud]

If this turns out to create headaches in the future, we can always export the data, pickle it, and commit it to our repo.

[sophiesheeline]

I think this makes sense -- we can be careful with how we use Airtable and (manually) add default values to and field that we don't yet have a need for. We can also delete ones like Intended Use or Notes if we don't think we will use them.

[rjbergerud]

Cool, I'll merge then.

For clarification, extra fields are all right. It's just that at some later point if code makes use of a field that we have to start checking and dealing with this cases.