OK - I’m going out on a limb. My focus for the last several years of my career was information management, but at a level pretty far above the technology layer. I was advocating for the management of information in much the same way a physical assets were managed (i.e., inventory, facilities, transportation, etc). What does the management structure of information look like? What are the RAAs (roles, authority & accountability).
I’m not and never have been a DBA. But, here’s what I think, with a fair degree of confidence that I’m correct. If you look back at an object DBMS and compare it to a relational DBMS it’s pretty much just a DBMS that has been optimized to run against a denormalized data model. I know, that’s an over-simplification, but it’s not that far off. Once you could stick a blob/clob into a relational column, a relational DBMS could pretty well handle it. The first iterations of Catia (very sophisticated CAD s/w from Dassault Systèmes) ran against a DB2 DBMS. I don’t know what’s the underlying data manager today, maybe it’s still DB2.
The No-SQL document paradigm is simply different. I don’t believe there’s anyway of building an effective and efficient No-SQL overlay on top of an RDBMS. Maybe someone who is deeper in the weeds with respect to the technology can correct my thinking, but that’s my perception. So if Oracle wants to compete in this arena (and they may not), they pretty much have to start from scratch, just like everyone else.