In 2003 Griffith University entered into a partnership with HarvestRoad to co-develop the Hive digital repository system. Issues relating to the base architecture (structure of collection bureaus, categories and sub-categories), security, permissions, integration with the LMS, granularity of objects to be managed, content reusability, base metadata standard, search facility, workflows, version control, content rendering and format, and timed release and content expiry were worked through. Policies dealing with intellectual property ownership, copyright management, and process ownership were developed. A number of start-up bureaus were established. This paper reports on progress, future applications, and lessons learned at Griffith.