Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign upReserve file-level DOIs? #7068
Reserve file-level DOIs? #7068
Comments
FWIW: I think reserving at DataCite, at file creation time (or manually upon request), would also work. (We'd have to make sure that deleting the file also deleted the reserved DOI at DataCite (allowed until the DOI is made findable) and that DataCite isn't charging for them.) There are two reasons that wasn't done already that I'm aware of:
That said, the main advantage of reserving with DataCite is to avoid other apps with the same authority/shoulder from reusing the same DOI which is unlikely at all given the naming schemes and even less likely if using the file DOI = / convention. So if DataCite want's to charge for reserved but not findable DOIs, it may not be worth it. |
adam3smith commentedJul 8, 2020
•
edited
At QDR, we're thinking of enabling reserving DOIs for files prior to publication.
User Story: As a researcher, I want to refer to a data file (think a specific interview or document) in a manuscript. I don't want to publish the data project until the manuscript is published and I want to avoid having to replace all file placeholders with DOIs prior to publication.
Implementation considerations: Given that files are often deleted and DOIs never published, we don't want to reserve those DOIs with datacite but rather reserve them internally, the way project DOIs are currently reserved in Dataverse.
Given the fact that files can be deleted (and, in fact, commonly are deleted and replaced in our curation process) we might want to consider making this available via a manual option rather than always (which may lead to misunderstandings and non-existing DOIs ending up in citations); on the other hand, auto-generating them is a cleaner solution with no additional UI requirement.
Question: Would this be something that IQSS would consider for the core dataverse code and if so, what are the main concerns that we'd need to address? (I didn't go back to look through the file DOI discussion to see what were the considerations there, so may just require a link).
CC @qqmyers