Over the past two weeks DSpace/Elements work has been divided between the micro and the macro: while the configuration of the systems’ technical back ends continues, big-picture discussions on the scope and role of Temple’s Institutional Repository (IR) are beginning in earnest.
Technical Developments
After finalizing technical preparations and preparing test versions of the Dspace->Elements and Elements -> Dspace crosswalk XML files, the Repository Connector was turned on, and the connection between Symplectic Elements to DSpace was tested. Unfortunately this revealed a few more issues that still need to be troubleshot. The SWORDv2 protocol by which Elements deposits files into DSpace isn’t communicating properly with Elements, making it impossible, as of right now, to send items from Elements to DSpace. Chin, Gabe, and Elements support are all working on the issue.
A little background: The connection between Elements and DSpace involves many moving parts. On the DSpace side three import/export systems for metadata and digital objects must be set up: a SWORDv2 service point, an OAI-PMH endpoint, and a RESTAPI endpoint. Elements harvests from DSpace via the OAI-PMH and REST endpoints, and deposits to it via the SWORDv2 protocol. On the Elements side, two crosswalk maps (documents written in XML that determine which Elements field values are mapped to which DSpace fields, and vice versa) must be manually configured. It is not just the systems that must be able to communicate, but also the servers on which they are hosted.
Big Picture Questions
While Gabe Galson and Chin Kim continue to hack away at these technical challenges, Annie Johnson, Holly Tomren, Cynthia Schwarz and Gabe Galson are preparing to tackle the big-picture questions surrounding Temple’s incipient IR:
How must our fledgling DSpace be structured so that it can accommodate the ever-expanding scope of our Elements implementation, as well as changes to Temple’s institutional structure?
How must our fledgling DSpace be structured to allow further development and accommodate additional affordances, such as the ability to generate DOIs through DSpace, or the ability to ingest dissertations directly from ProQuest?
How should our DSpace be structured so as to best serve the end user, and to promote discovery?
How can we make sure our IR metadata retains its utility through (and beyond the life of the systems that currently host it?
And many more!
Next week and the week after Gabe will schedule brainstorming meetings with @anniejohnson and @hollytomren, then follow up with formal stakeholder meetings to formally set the course of this project. Holly’s metadata expertise and DSpace experience will help us structure our DSpace and XML crosswalks in a sustainable and scalable way. Annie’s knowledge of institutional repositories, scholarly communications issues, and Temple’s own scholarly community will allow us to deliver a product that satisfies this community's needs. On Annie's suggestion, the Temple team will get in touch with Paolo Mangiafico at duke, who lead the team responsible for the wonderful DukeSpace DSpace implementation, a project centered around a DSpace-Elements integration undertaken in-house. Stay tuned!