Closed illip closed 3 years ago
I +1 this. CIDOC CRM puts in the primitive data types as placeholders so as not to force the adopting institution to use particular kinds of data type. When it comes to standardised implementation, however, one should choose a consistent data type and document this.
You do not also use rdfs:literal?
@Habennin, I don't know if I understood your question about rdfs:Literal. In fact, our XSD datatypes are used precisely on rdfs:Literal as attributes. I don't know where we could apply them otherwise. :p Perhaps, I missed something!
Currently, TM2.1 is not consistent in the use of XSD datatypes over primitive values. CHIN has decided to edit the documentation to remove all primitive value mentions and this will be done in the 2.1 version.
All the primitive value mentions have been removed. We also replaced all the xsd:date
by xsd:dateTime
and edit the literals for yyyy-mm-ddT00:00:00
.
Currently in our Target Model 2.1 (will be released soon), we mention CRM's primitive values (e.g.
E62_String
) and XSD types (e.g.xsd:dateTime
). I would recommend to harmonize our documentation by using only one representation.Following this CRM's encoding rule:
I think we should remove CRM's primitive values to avoid confusion.
As far as I know, currently we just use
xsd:string
andxsd:dateTime
.What do you think?