Discussion - Schema usage

[dogversioning]

The spirit of this ticket is to hash out a direction on having a heavier namespacing at the schema level.

Some points discussed:

• Pro-higher granularity schemas:
  • Less verbose namespacing inside of a study
  • easier to archive at the study level
• Pro-lower granuality schemas:
  • Less infrastructure overhead (though we can automate this with the right permissions)
  • easier to archive everything at once

Making a change to schema handling would be a large breaking change, so if we do it, we should allot appropriate time for it.

[dogversioning]

Per #252, we have a mechanism to selectively specify in a manifest that a study can be in its own schema, and the use case is limited. More information about the rationale for this is in the MAINTAINERS.md file.