Record class, schema maintenance and double work

[mguidoti]

Hi guys,

So, I noticed the Record class in the intial PR by @ctrl-alt-soneca and I checked Zenodo Schema Deposit schema and I was left wondering again about governance and maintenance of frictionless schemas, and if it's not double work to use classes for records instead of native data structures (dicts), since we would have to declare a whole bunch of attributes and maintain that and the schemas individually if anything changes, to the best of my knowledge.

What are your thoughts, again, on these three separate issues?

One fourth and additional question is regarding custom fields. In order to add those to a fricitonless schema, we would most likely have to list them all. Do you guys have the list of all fields added to this point? What are your thoughts on this too?