mutliple PID for affiliation (ROR, RINGGOLD, ...)

[mellybelly]

https://ror.org/

We have found great inconsistencies across different venues for contributor orgs. ROR is not perfect but it helps ;-).

[jcolomb]

We probably need to find a way to give multiple PID for both organisation and people.

In particular, orcid is not using ROR, is it?

[mellybelly]

Agree, the model should provision for 1...n person PIDs, and perhaps 0...n org PIDs.

ORCID was planning to use ROR and has been a contributor and partner in design; conversations have been going on for >5 years so I'm hoping recent advances in tech and governance for ROR will push us over the edge!

[jcolomb]

in #17, we have (following jats4r)

PID:
      type: RINGGOLD
      id: 9373

Will it allow multiple elements by having multiple PID elements, or should we nest multiple elements in the PID entry ?

like this (note both PID works for HU Berlin):

PID:
      type: RINGGOLD
      id: 9373
PID:
      type: ROR
      id: 01hcx6992

or

PID:
  -
    type: RINGGOLD
    id: 9373
  -
    type: ROR
    id: 01hcx6992

[nathanlesage]

I think we are seeing some overlaps with regard to the ID systems, right? I'll just drop my idea in here for completion, but we might want to center the author and affiliation-id discussion around one single issue…?

affiliation:
  - name: institution name
    ringgold-id: asdf
    ror-id: wasd

This way the parser could just save all available IDs, if there are any, which would solve them being optional (if no ID-properties are defined, then there is no ID), as well as stackable (all available IDs can be stored internally in the parser using any arbitrary AST)

[jcolomb]

indeed solved in #20