Open andrea-perego opened 5 years ago
My understanding was that we had listed only terms which have normative status, namely old and new DCAT terms, and terms borrowed from other w3c recommended vocabularies (e.g., PROV, ODRL, SKOS). dct and foaf are two exceptions, due to the fact that they are long-standing "standard".
Have I misunderstood it?
ADMS is a W3C Group Note and I am not sure if recommending vocabulary terms borrowed from a W3C working note is acceptable for a W3C Recommendation. This could be one possible reason why we have ADMS in Guideline examples but not ADMS in the vocabulary specification which is normative.
I remember the concerns about relying on non-normative vocabularies when defining the Data web best practices. The assumption at the time was that a recommendation shouldn't recommend using something that has the working note status.
I wonder if we should check this with our W3C contact point.
This does not answer the question, but the ADMS vocabulary is online here: https://www.w3.org/ns/adms#
Note that DC Terms is in the process of becoming an official ISO standard, which should help folks who are tasked to favor ISO standards. It will be iso15836-2. I'll send more info when I get it.
@riccardoAlbertoni , +1 to check with @draggett .
However, my take is that we are not prevented from using "officially standard" vocabularies, provided that there's a versioning policy in place ensuring that the vocabulary won't disappear, and it won't be changed without taking into account existing implementations.
Although ADMS is not a Recommendation, the W3C preservation policy prevents it from disappearing or even changing URL or namespace URI. Moreover, changing the vocabulary itself would require chartering a new Working Group, who should take into account existing implementations, etc.
So, I think we can safely re-use ADMS terms in normative sections - and they same should apply for DQV and DUV.
To be decided if this should go to milestone DCAT Future Priority Work.
@andrea-perego: Do you still think we need to discuss this issue? Or we can keep the distinction between Normative and non-normative namespaces as a driving principle for deciding what we need to include in the properties section?
adms:identifier
is present only in the guidance section on identifiers.I guess it should be at least added to the section describing class
dcat:Resource
as one of its properties.