This is the only property whose range is set using rdfs:range.
For every other property the range expectation is set using schema:rangeIncludes, which is consistent with the weak axiomatization pattern agreed to for the SOSA graph.
However, the expected range for schema:rangeIncludes is a Class so this probably should not be used for xsd:dateTime.
It is also effectively the only place in SOSA where a specific encoding instruction is introduced into what otherwise can be understood to be a 'conceptual model'. Encoding instructions may be better managed in a different layer - e.g. SHACL rules implementing specific profiles.
I wonder if this axiom should be removed from the SOSA graph?
The range (expected value) of
sosa:resultTime
has always been an encoded date-time:xsd:dateTime
. This is expressed in SOSA asThis is the only property whose range is set using
rdfs:range
.For every other property the range expectation is set using
schema:rangeIncludes
, which is consistent with the weak axiomatization pattern agreed to for the SOSA graph. However, the expected range forschema:rangeIncludes
is a Class so this probably should not be used forxsd:dateTime
.It is also effectively the only place in SOSA where a specific encoding instruction is introduced into what otherwise can be understood to be a 'conceptual model'. Encoding instructions may be better managed in a different layer - e.g. SHACL rules implementing specific profiles.
I wonder if this axiom should be removed from the SOSA graph?