Great thought-provoking post https://lnkd.in/eb4aWWh6 from Ludo Waltman and André Brasil on the challenges of assigning DOIs, and metadata more generally, for innovative types of publishing. This really reflects my own challenges working with metadata of this type at Research Consulting Limited. In a previous role overseeing the DataCite account for an institution, I came across similar challenges in deciding when to assign DOIs. Institutional repositories regularly hold versions of content which have been assigned DOIs elsewhere. I ended up concluding we should become okay with giving multiple DOIs to different versions of content (even if different meant "hosted elsewhere"). In avoiding assigning a DOI because one already exists, we compromise the discoverability of valuable content and create a two-tier system of content. Plus there is already good metadata available within the DOI system to indicate where items are versions of each other. However, to capitalise on this data use of versioning metadata needs to be consistent and reliable. Bibliometric systems need to be able to automatically identify alternate versions and allow records to be collapsed together for analysis. I'm not certain whether the challenge in doing this lies more in the productionn of the original DOI metadata or how it is used in bibliometric systems, but I would be really interested to hear from those with insights. Final thought: I'm personally a fan of when one organisation mints DOIs for multiple versions (i.e. all records have the same DOI prefix) and indicates version information within the DOI itself. If it was possible to make the existence of an alternative version transparent in DOIs regardless of which organisation minted it, that would would be my dream! But infinitely easier said than done, I realise. #scholarlycommunication #persistentidentifiers #PIDs