What’s the point of having open scholarly infrastructures and how do we test their resilience? | Martin Paul Eve | Professor of Literature, Technology and Publishing

"For me, the fundamental meta-principle, or ideal, that underpins POSI (the Principles of Open Scholarly Infrastructure) is forkability and persistence. Taken on aggregate and implemented, an organization that signs up for POSI should be duplicable. That is: I should be able, as a reasonably technically competent individual, to acquire all the components of a POSI-posse signatory, and rebuild/resurrect their technical architecture....

But why do we want this “forkability”? The simple answer, as with many things at Crossref, where I work, is: persistence. The scholarly infrastructures on which we depend for (say) linking and metadata resolution must continue to exist and there must be mechanisms for their ongoing operation. POSI is designed to allow an arbitrary third-party entity to continue the operation of an infrastructure in the event of the original organization’s demise – or in cases where the original organization might be acting against the wishes of the community....

A point that disturbs me, though, is that we have never tested this forkability on any of the key scholarly infrastructures. Wargaming simulated disasters is common practice in organizations. At the Open Library of Humanities, we simulated a total systems failure and resurrected everything from backup. Crossref has, in the past, found that even massively redundant systems can go wrong if not tested...."