I went to a most useful talk this morning by Stephen Carlton (@LivUniOA) on the Univeristy repository. I had whinged about this as being nearly unusable, but then I jumped in on an early version.
The repository is now useable, though it is quirky. A few lessons from my efforts to update my entries.
Something that I don’t quite follow yet, is that I get an error message relating to my id. I guess this is may be due to current university systems expecting my name in one format, but when citation information comes down from PubMed or CrossRef (I used both, seem equivalent) our system cannot parse the different forms of a name that appear: David G Fernig, DG Fernig, D.G. Fernig etc. Alternatively I have done something wrong, always a good place to put your money.
[update #1: it looks like it cannot parse, you have to remove author info, remember to cut surname. paste this back in and it comes up with the goods]
Some data have to be entered manually, largely radio button options, though in some instances there is a need to paste in an abstract. Slightly frustrating, though there is a whole page for data entry, the system only seems to be able to cope with one or two items at a time, so click, save, return, click, save, return, paste, save, return and so on. When I input all the information in one go, then either hit ‘save’ or hit ‘next’, I was then informed I hadn’t entered some of the information. Returning to the relevant page indeed showed an empty field.
Otherwise it works, though it isn’t ideal. It would benefit from using Orcid id, since this allows for batch processing. However, for current publications I am not sure this would reduce the amount of information we have to input. These will always be on a one-at-a-time basis, since we have to comply with HEFCE REF rules relating to the time allowed between acceptance and deposit in the repository. Documents still have to be uploaded and it isn’t entirely obvious to me that issues arising from publications authored by multiple members of Univeristy staff will be simply aggregated into one record.
The final lesson. Publish OA, in OA journals or learned Society journals. The commercial sector makes uploading a document a pain, because it has to be a submitted version of the paper and you have to find which one complies with copyright restrictions. Each publisher has different rules. By going OA, you halve (at least) the time required to deposit articles, as well as having the beautiful typeset version of your work in the repository, rather than some Word document, with figures at the end.
Ferniglab Blog 