User:TiagoLubiana/Colons, underscores or numeric only
When adding properties for OBO Foundry IDs on Wikidata, there is one big decision regards formatting: should I use IDs with colons (:), underscores (_) or just leave the numeric part?
This page is an attempt to list some of the arguments there.
E.g.
- Numeric only: http://purl.obolibrary.org/obo/PO_0009072
- Colons: PO:0009072
- Underscores: PO_0009072
Numeric only
[edit]Charlie Hoyt, from bioregistry.io and other registry people (e.g. identifiers.org) don't like duplicating the information about the source, as it causes a "banana problem" (i.e., a redundant copy of the name/acronym of the database in the local unique identifier). It is already in the Wikidata property, so why should it be repeated in the value?
Colons OR underscores
[edit]Having colons or underscores indicating the property makes it possible for the ElasticSearch engine to find if someone does a search with the prefix.
Colons
[edit]It is the canonical way of formatting CURIEs. https://www.w3.org/TR/2010/NOTE-curie-20101216/
Underscores
[edit]It makes it easy to build an URL from the identifier, e.g. http://purl.obolibrary.org/obo/PO_0009072