@ignoramous What do you think about PG integrating a simple signed timestamp authority (RFC 3161)? While this would only prove _when_ data existed, not _whether it was correct_ , it seems more attainable in the near term while still working toward the same goal.