Closed cdmgtri closed 1 year ago
I think it's going to be important to consider who this will affect.
NIEM XML is our biggest user base by far. This would not be popular:
<my:Message>
<nc:PersonGivenName>
<nc:PersonGivenNameLiteral>John</nc:PersonGivenNameLiteral>
<xml:lang>en-US</xml:lang>
</nc:PersonGivenName>
<nc:PersonSurName>
<nc:PersonSurNameLiteral>Smith</nc:PersonSurNameLiteral>
<xml:lang>en-US</xml:lang>
</nc:PersonSurName>
</my:Message>
when the following is easily available:
<my:Message>
<nc:PersonGivenName xml:lang="en-US">John</nc:PersonGivenName>
<nc:PersonSurName xml:lang="en-US">Smith</nc:PersonSurName>
</my:Message>
Currently, NIEM JSON has a much smaller user base, and representing NIEM attributes as a new rdf:value
or a named literal property only when needed seems to be a good compromise.
There are even fewer users right now for NIEM RDF. I'd be very reluctant to make a big breaking change in NIEM 6 for XML in order to make the RDF cleaner.
Also, RDF-star and JSON-LD-star indicate that others have found it useful for being about to make statements about statements (annotate objects and literals) in a consistent manner.
We can't take the attributes out of NIEM XML messages. That was a crazy-Dr-Scott idea from a while back. I'd still like to gently discourage semantic attributes, but if folks want them, they get them. I suggest we close this issue.
Things like nc:PersonGivenNameLiteral
do show up in CMF. I think we need them in NIEM RDF, probably in NIEM JSON.
BTW, the xml:lang
attribute needs special handling. Take a look at the last section in CMF and RDF representation in NIEM 6.
BTW, the
xml:lang
attribute needs special handling. Take a look at the last section in CMF and RDF representation in NIEM 6.
Updated link.
Migrated from NIEM/NTAC#56
Change all the attributes containing semantic data into elements, leaving just the infrastructure support stuff.
Christina: Some attributes, like text string stuff would require text elements to have a inner element.
Dr. Scott: Choices