On Sun, 30 Nov 2014 07:47:46 -0500 Jonathan A Rees rees@mumble.net wrote:
I'm very skeptical of applying 2119 to vocabulary specifications. I don't think there's any clear agreement on what constitutes conformance to a vocabulary specification
On the level of specifying the meaning of a vocabulary, I think I agree with you. It doesn't feel like specifying that dwc:scientificName MUST/SHOULD/MAY carry something that is actually a scientific name is particularly helpful.
However, the DarwinCore RDF guide feels like an implementer's guide in how to write applications that produce and consume RDF using TDWG DarwinCore vocabulary, a specification that, in order to be understood and not cause problems for consuming applications, producers of DarwinCore in RDF should e.g. assert dwc:recordedBy "Asa Gray" or dwciri:recordedBy http://viaf.org/viaf/7504476 but not dwc:recordedBy http://viaf.org/viaf/7504476 or dwciri:recordedBy "Asa Gray".
http://www.w3.org/TR/EARL10-Guide/ feels particularly relevant - it is a developer's guide that accompanies the EARL vocabulary specification http://www.w3.org/TR/EARL10/. In an earlier draft of EARL, the language of RFC 2119 was explicitly invoked http://www.w3.org/TR/2009/WD-EARL10-Schema-20091029/, in the latest version, the RFC 2119 language was removed from the vocabulary specification and relegated to the develper's guide. This feels very paralell to TDWG DarwinCore and the TDWG DarwinCore RDF guide.
-Paul