Fedora is the underlying architecture for a digital repository, and is not a complete management, indexing, discovery, and delivery application.

It is a modular architecture built on the principle that interoperability and extensibility are best achieved by the integration of data, interfaces, and mechanisms (i.e., executable programs) as clearly defined modules.

[7] Cornell and UVA formally disputed the request and, as a final settlement, all parties settled on a co-existence agreement that stated that the Cornell-UVA project could use the name when clearly associated with open source software for digital object repository systems and that Red Hat could use the name when it was clearly associated with open source computer operating systems.

The objects contain linkages between datastreams (internally managed or external content files), metadata (inline or external), system metadata (including a PID – persistent identifier – which is unique to the repository), and behaviors that are themselves code objects that provide bindings or links to disseminators (software processes that can be used with the datastreams).

Out-of-the-box Fedora includes the necessary software tools to ingest, manage, and provide basic delivery of objects with few or no custom disseminators, or can be used as a backend to a more monolithic user interface.