search

nfdi4plants / isa-ro-crate-profile

MIT License
4 stars 0 forks source link

Set family name from MUST to SHOULD #15

Closed HLWeil closed 2 months ago

HLWeil commented 5 months ago

https://www.quora.com/What-cultures-do-not-use-family-names#:~:text=As%20well%20as%20the%20mentioned,Sometimes%20they%20invent%20a%20surname.

Some bigger culture groups like javanese people from indonesia don't have surnames, so we should not set this field to MUST.

Either use MAY or SHOULD instead.

floWetzels commented 4 months ago

Sounds reasonable. But how about the first name? Should there be any MUSTs in the person profile other than the ID?

HLWeil commented 4 months ago

Hmm, good question. Probably also cases where an ORCID does suffice. So yeah maybe set it to SHOULD.

stuzart commented 3 months ago

I completely agree about the cultural differences, and shouldn't be a MUST. Could just replace both with name instead, this is what the draft bioschemas profile recommends: https://bioschemas.org/profiles/Person/0.3-DRAFT

floWetzels commented 3 months ago

Good catch! So should we say name is a MUST, but familyName and givenName are COULD? Or drop them completely?

stuzart commented 3 months ago

I would say keep name as MUST and drop familyName and givenName

floWetzels commented 3 months ago

Ok, do you agree, @HLWeil @Freymaurer ?

HLWeil commented 3 months ago

I agree that this would be the most practictal solution with no further constraints. But as we want to depict the ISA model with this profile, I think we should stick to givenName + familyName, but just loosen the restriction of familyName.

stuzart commented 3 months ago

sounds good to me, as you say we're constrained by the ISA model.

floWetzels commented 2 months ago

Fixed by PR https://github.com/nfdi4plants/isa-ro-crate-profile/pull/19