I think we need to recognise different breeds of repositories. An institutional repository of OA research papers has distinct characteristics from a national subject-based repository. In order to be useful I think the definition needs to make reference to different types whilst pointing out similarities. if the definition is useful people need to see it doesn't just apply to vanilla implementations: their repository with all its particular functionality and interactions with other systems counts too.
Amber, I think you are right, but I read this as relating to assumptions about implementation architecture, not target market. A data repository might have different requirements than an OA text repository, and we should take those into account. But it's less important how the implementation is actually designed; filestore plus catalogue might be perfectly adequate implementation.
Revise
Vote Up
Vote Down
Post New Comment
Close