Warning: These documents are under active
development and subject to change (version 2.1.0-beta).
The latest release documents are at:
https://purl.dataone.org/architecture
Although content in DataONE is immutable, there are situations where the original content can be updated, and in these cases a new copy of the content is created for science metadata. The update process for data is currently not defined, though for practical reasons it is unlikely to be feasible to preserve all copies of data, especially for particularly large data sets.
The premise is that owners of the derived products will be notified - for this to occur, the relationships between original and derived need to be recorded and discoverable. The mechanism for this requires further design. It would be possible to record such relationships in the system metadata, or perhaps it may be necessary for a different class of object to be created to specifically to support recordation of relationships between content.
Notes
Figure 1. Interactions for use case 28. Versioning – Derived products should be linked to source objects so that notifications can be made to users of derived products when source products change