One of the most important challenges we are facing as a repository is embedding our repository within other institutional systems. This is made a little more tricky as we are a consortium of three with a single installation of software at one partner site. Should deposit be invisible to the depositor? Should our capture processes be so subtle and moved so far towards the author that, effectively, they need to take no additional action for the work to be captured and deposited? Moving towards "desktop deposit" of "desktop capture" may indeed be a desirable goal. This may be particularly key if the process of metadata creation becomes detatched from the capture of the relevant version of the file. For example, we are now seeing an emphasis on bulk capture and import of metadata. University of Leeds has recently purchased the Symplectic Publication Management System which merrily trawls PMC and Web of Knowledge for Leeds publications, pulls in the metadata and uses this to populate a University publication database. We are told that we'll also be able to take a feed of this metadata into the repository and make this openly available. Great!
The potential drawback here, though, is that the emphasis on post publication, post indexing metadata capture means a long (potentially very long) delay between work being accepted for publication and metadata appearing in the local publication database and institutional repository. Asked when they would like to deposit their work, responses from White Rose authors varied but the most popular deposit point was "at or after publication". There is a danger that collection of metadata and compilation of publication lists is stongly associated in the researchers' minds with administrative and accounting processes. There is a danger that collection metadata and publications within repositories is also seen as a largely bureacratic, adminsitrative process. This, in turn, leads to the potential danger of seeing deposit as "a summer job" or some kind of periodic, mopping up process.
Clearly, if we are to capture the research we aim to, we need to have efficient, relatively pain free processes to capture research at the relevant point. This is likely to be at the point it is accepted for publication. This means that the process is less "tidy" because we won't, at that point, have complete publication metadata for the work. We need to consider what key metadata we can capture automatically and what information needs to be supplied by the author.
If the repository is effectively "invisible" to the author and there is no explicit process of deposit we need to think carefully about what this means for advocacy. Authors should be aware of the key aims of capture - not for bean counting but to foster dissemination, visibility and impact, reuse of research.
White Rose Research Online will continue to work towards better embedding within local systems. An initial step will be the proposed linkage between WRRO and Symplectic in Leeds but we will also be keeping a close eye on developments at Sheffield and York - in particular, any systems which are put in place to handle REF submission.