TU Delft Library recently held the first brainstorm meeting to rethink the Library Repository.
While the initial theme of the discussion was the repository, the conversation quickly moved away from the repository and focussed more on the content contained within them.
To think about a repository infrastructure first needs thinking about (digital) collections. What collections does the library have? And how are they stored, managed and published? And what collections does the library want to ingest in the future? And how will they be stored, managed and published?
Only by answering those questions will you develop the requirements needed to shape a new repository(s) infrastructure.
This conversation then broke down into more specific questions about collections
Defining the scope of a collection.
Legacy thinking about print collections sometimes unduly influenced discussions about digital collections. For example, reading lists have traditionally not been part of a library collection. But now university teachers are increasingly publishing reading lists, but via a wide variety of tools all over the web. Could the library have a role in collecting and publishing reading lists?
Equally, there has never been a specific approach (at least at TU Delft) to collecting reports, working papers, strategy documents (the so-called grey literature). Or outputs such as posters and presentations. Could this be changed if supported by a new approach to this type of digital content and a new infrastructure for managing and disseminating them?
The areas where there was the significant lack of clarity concerned the ‘research collection’. Here, the divergence between a traditional print collection for research and a digital collection is the greatest. How should a library approach managing the research collection of the future? Particularly for the core unit of scholarly communication, the research article.
With the increasing frequency of articles disseminated by published as Gold Open Access, is there a point to the library publishing earlier versions of the article in an institutional repository (which have less worth as they are not considered the final peer-reviewed version)?
Indeed, could a library think of its research collection in terms of simply being a metadata collection? Ie, leave the management and publication of research outputs to third parties, and simply help access by maintaining high-quality metadata that links to these research outputs? Or does such an approach abdicate the library’s historic task to safely archive knowledge in the long term?
Governance of collections.
There was also discussion over the governance of collections. Sometimes, the collection development may have been driven by another party within the university – another service or academic faculty – who then worked with the library informally. What happens when that party leaves, or simply shows less interest in the collection? Does the library have the right to take ownership of what happens to others’ collections?
How can the demands of a Collections Strategy be turned into technical requirements?
The conversation finally turned back to how to connect the demands of digital collection curation into requirements that could underpin the development of new or updated technical infrastructures.
Initially, it would be impossible to set down a complete list of requirements. But it is possible to start with some high-level requirements (eg., a specific digital collection needs management of high-quality metadata, DOI per digital object (or even a part of object), user upload and management of his / her own files, quick search mechanism, high-quality browse function, statistics on document usage / downloading)
Such requirements can then be matched again potential infrastructures to support the new repository.