
From the Data Model WG charter:
[The WG] provide[s] a framework for the description of metadata attached to observed or simulated data.
[...] focuses on logical relationships between these metadata, examines how an astronomer wants to retrieve, process and interpret astronomical data, and provides an architecture to handle them.
[...] What is defined in this WG can then be re-used in the protocols defined by the DAL WG or in VO aware applications.
<dataType>
<vodml-id>source.SkyCoordinate</vodml-id>
<name>SkyCoordinate</name>
<description>...</description>
<attribute>
<vodml-id>source.SkyCoordinate.longitude</vodml-id>
<name>longitude</name>
<description>...</description>
<datatype>
<vodml-ref>ivoa:quantity.RealQuantity</vodml-ref>
</datatype>
<multiplicity>1</multiplicity>
</attribute>
...
src:source.SkyCoordinate.longitude
points to
<attribute>
<vodml-id>source.SkyCoordinate.longitude</vodml-id>
<name>longitude</name>
<description>...</description>
<datatype>
<vodml-ref>ivoa:quantity.RealQuantity</vodml-ref>
</datatype>
<multiplicity>1</multiplicity>
</attribute>
Can be used in VOTable (pending VOTable discussion)
<VODML>
<ROLE>src:source.SkyCoordinate.longitude</ROLE>
</VODML>
utype -> vodml-refDocument seems stable. Should be ready for Proposed Recommendation pending few TBDs.
Implements UML concepts of redefines and subsets
subsets: WG chairs are a subset of the IVOA members
redefines: the SphericalError on a SphericalCoordinate is a subtype of a generic Error on a Coordinate
STC2 uncovered issue with subsets property
The subsetted property would redefine its VODML-ID, breaking inheritance, e.g.:
<ROLE>stc:SphericalCoordinate.error</ROLE>
vs
<ROLE>stc:Coordinate.error</ROLE>
Fixed by making subsets an attribute of the super-type.
Formal meta-modeling rules make models more robust and interoperable. We need more tools for easily translate models to VODML/XML.
stc2, char, spec2)or
Consensus seems to be that they are globally unique.
Can we confirm this?
Should models (and prefixes) be registered?