Page Comparison

ATTENDEES: Wendy, Jon, Darren, Oliver, Dan S.

Annual Report

No additional comments, will leave available through Sunday and then send in

Codebook

• Field level documentation HTML version has been completed

• Wendy will review Change Log and update, set up review page and create the google page for the high level documentation draft

• Review should start by end of May. This should last to the middle or end of August

• Anticipate easy review because there has been a lot of consultation - this should mean that developers can begin work using the Review version we some certainty

COGS work

Reference:

• Reference properties outside of the identification is the issue

• Reference is an entity and will act differently with serialization - RDF is a URI, XML / JSON model

• Review specifics of properties - Wendy will document

  • Example: Where you need to know a context of what you are referencing add another reference for context.

  • Including extra properties

• Option is to create Complex types

  • Reference becomes an element in the complex

  • Where is such extended content needed

Identifiers:

• don't think there is any difficulty by serialization. Some have their own identification (URI in RDF)

• Drop the difference between Versionable and Maintainable. There is only one way to identify an item (URN, AGENCY, ID, VERSION)

• Additional properties

  • Describable class extension of versionable

  • Properties found in versionable - review

• Identification is injected during creation of serializations

  • Inject identfication into any item type will have identification

  • Complex data types DON'T get Identification

• Another area of confusion for additional properties is the use of Name/Type/Description and Citation (currently used singularly or together in some elements)

  • Look at where citation is used and how. OtherMaterialTypes (clean-up later) Group? StudyUnit? ResourcePackage, Instance, etc.

CHOICE issues are being sorted

PhysicalDataProduct

• Add NCubeInstance to the basic RecordLayout now in the main schema

• Make the others derivatives of the BaseRecordLayoutType with unique element names and move the content to PhysicalDataProduct

• This will get everything in AND let us use one PhysicalDataProduct to import

Representations

• Recreate the few representation types that are not in the ManagedRepresentations.

• This gives them the same structure and we can determine later if we want to continue this division between representations that have references to other objects or what.

CV and RDF resolver:
The sandbox will be set up end of next week

https://rdf-vocabulary.ddialliance.org/cv
e.g. https://rdftest-vocabulary.ddialliance.org/cv/ModeOfCollection/1.0.1