ADMIRAL 20100825 technical meeting with library services
From ImageWeb
Contents |
Technical Meeting with Library Services
When: 25-Aug-2010, 11:15-13:00
Where: Zoology, room B-17
Present: AR, BW, GK
Goals
Meeting with AR about databank progress
- review progress since last meeting (17 Jul?)
- articulate and review next steps for Databank development
- general technical discussion and review
Progress review
The main progress since our last contact has been the implementation of RDF data merging from the manifest supplied with a dataset. We ran the test case previously written, and exposed two problems:
- the manifest.rdf file must be at the very top level of the ZIP file, rather than in a single directory contained in the ZIP file. AR may allow the latter as an alternative case, as it is sometimes easier to create ZIP files containing a single directory.
- when creating a dataset from a ZIP file (as opposed to adding it separately), the owl:sameAs for an alternative dataset URI is not generated.
Other improvements include fixing some link URIs in web pages served, and updated API documentation.
AR and NJ have agreed a feature set for V1.0 of RDF Databank, but this is not yet document. From informal discussion, the features appear to be a good match for anticipated ADMIRAL requirements. Timing for V1.0 deployment is not yet clear, but other research groups are now showing interest.
Next steps
Next implementation steps for Databank are:
- State information (version list, etc.) to be served at a separate URI from manifest and metadata information, rather than using content negotiation for this.
- return JSON or RDF versions of metadata and state information.
- Versioning: new versions to be generated for all update actions; default access most recent version, but append version to URI to retrieve older versions.
GK/BW expect to start working on tools for real research data submission some time after 7 September.
