As advertised, the objective of this breakout was to review the current and planned prototypes and related activities, including a related ISO group, look for overlaps and gaps, and formulate plans for future work. Larry Lannom gave an overview of work to date. Steve Richard, attending remotely, described the use cases he was covering and his progress on a JSON schema, testing with Cordra (registry software), and plans on moving model compilations from spreadsheets or RDF into Cordra. Our second and final remote presentation was from Mike Finnegan of the Vermont Monitoring Cooperative. He described their fairly extensive use of types, their plans to expand use of types from the attribute to the data set level, and their work with CNRI to use types in a dataset recommendation system. Tobias Weigel described the role of types in climate data processing, across different projects and different use cases. Ulrich Schwardmann reported on the use of DTR at ePIC, including use of type hierarchies, a defined type life cycle, and implementing the International System of Units in a DTR. This generated the most lively conversation of the breakout, around the issue of representing these units in a type registry and the representation of one unit by another. Our final project report was from Andres Ferreyra on AgGateway’s Progress on Data Type Registries. He described the ContextItem system developed by the AgGateway consortium and its connections with the RDA DTR WG. Larry Lannom finished the meeting with an update on the ISO Study Group on data type records. The group is scheduled for completion in June of this year and will likely either create a technical report or move to create a new ISO standard. Input to the study group was derived from members of this RDA WG. The results of this will clearly affect future work, which will revolve around increasing sophistication of prototypes and the upcoming need to connect them and to figure out where base types reside and who is responsible for them.
The presentations are linked below: