Closed danfowler closed 7 years ago
Except then anyone interpreting the file as CSV would also have to understand how to read and understand JSON, at which point why wouldn't they simply use the JSON dump? Pipe (or semicolon) delimited fields in CSV are very common and understood by our users.
Yup! I was half intending to say... "or not!" given how common a use case that is. In that case, should we add an issue for the specs: https://github.com/frictionlessdata/specs/issues
Thanks for the suggestion, but I don't think we'll implement this at this time.
Lists of items in a field can be expressed as a JSON array instead:
http://specs.frictionlessdata.io/table-schema/#properties