Presentation of use cases and mock-ups for persistent identifier versioning.
Background
The most common support request on Zenodo is related to modifications of files which essentially falls in three categories:
- Update of data - e.g. new data was collected, processing of data was improved, peer-reviewers made comments.
- Mistakes - e.g. cosmetics such as typos or accidental inclusion of hidden files and folders in ZIP files, renaming of files, accidental inclusion of confidential data or mistakes in exported data formats.
- Deaccession of data - e.g. copyright infringements, breach of confidentiality, data collected from human subjects or scientific invalid data.
Once a persistent identifier (PID) such as a DOI or an EPIC is registered, the underlying data files should no longer be modified. This is to ensure reproducibility and specificity - i.e. when a resource such as a publication or dataset is being cited using the PID, it should be possible to resolve the PID in order to gain access to the specific resource which was cited, and not some later modified version of it. See https://doi.org/10.7717/peerj-cs.1 for more information.
Both B2SHARE and Zenodo currently have no way to let users autonomously update their uploaded resources while both do have administrator facilitated ways to handle both deaccession of data and cosmetic mistakes in files.
Scope
The primary scope of persistent identifier versioning is to allow users to autonomously update files for their uploaded resources while maintaining a trusted repository operation.
