Definition

Flow-based programming (FBP) is a Dataflow programming paradigm with a unique set of characteristics on one hand and a kind of Component-based software engineering approach on the other hand. It was created by J Paul Morrison while working for IBM in 1970s.

Flow-based programming considers an application as a set of components ("black boxes") and connections between them. In other words, application is a graph consisting of components linked by connections. Each component in the graph can also be seen as a "white box", which is either a nested graph (subgraph) or a subprogram in a conventional programming language (usually textual). The approach is mostly focused on on data flowing in form of Information Packets (IPs) between the nodes, rather than control flow in particular subprograms.

Characteristics of Flow-based programming

FBP is often confused with Dataflow programming, but in fact it is a subset of Dataflow programming with the following unique properties:

• Structural: graphs are structured, components have structure too (interface, state and behavior).
• System design is split into 2 layers: graph layer (usually visual) and component layer (usually textual).
• Parallel: each component runs in its own process, thread, coroutine or other concurrency facility.
• Asynchronous activation:
• Data is encouraged to be immutable, but it is not a must. Information packets can be mutable, each IP has its lifecycle and is owned explicitly by one process at a time.
• Components can be stateful.
• Components can have multiple inputs or outputs.
• Application is a graph rather than a tree. Cyclic connections (feedback loops) are allowed.
• Connections are implemented as bounded buffers with FIFO order and capacity from 0 to a number limited by implementation.

See Concepts for further details.