Open eteq opened 4 years ago
I don't rightly remember, it could be because the original data had numbers for the references. Key should be unique and used when possible. I consider references.json to still be a prototype implementation.
Gotcha. Going forward I'd say it makes sense to use only key
then - I think the original data reference numbers were just there as because it made sense for the table format so we can probably let them go once things have proceeded past prototype.
I noticed the example references file (https://github.com/dr-rodriguez/DSII_LocalGroupDB/blob/master/references.json) has both a 'key' and an 'id' field. What's the distinction there? Is one not necessarily unique? If not maybe just 'key' is necessary? (Or "id" can be auto-generated somehow at runtime if it's a db thing?)