Well, in this case that's one of the things that makes it an ORDBMS engine.
One of the interesting things about writing the series I have been working on PostgreSQL as an ORDBMS is that it has helped me solidify my understanding of how these features fit together. It has also forced me to think about separation of concerns a lot more.
What this has taught me so far is that PostgreSQL is an absolutely amazing data modelling platform. Sure that's a lot more than an RDBMS engine. The difference is somewhere between a math program capable of doing algebra and one capable of doing symbolic manipulation to solve derivatives and integrals. (Indeed this is forcing me to rethink the way I approach relational math to account for the sorts of things an ORDBMS can do.) So yeah it is a lot more.
But "just a database" is a hard statement to nail down. If BDB is "just a database" then MySQL is "not just a database." As Paul Brown put it in "Object Relational Database Development: A Plumber's Guide" this basically makes an ORDBMS an information backplane for software services. It becomes an interesting whether whether "database" is a superset of that or not.
when your code goes into RDBMS engine, then it's no longer just a rdbms engine