The flow editor displays the flow path of a flow as a series of various-colored icons (flow nodes) connected by black lines capped with arrows (connections) as illustrated in Figure 1 following. The arrows indicate the direction of execution:

For a description of the nodes in the flow path shown in Figure 1 see Contents of a flow.
The flow path of a flow can contain alternate branches (flow branches) and loops.
You can create flow branches in the following ways:
In a nonterminal flow (see Nonterminal flow) you can create flow branches by using a Switch node with multiple output terminals, each output terminal connected to another flow node in the flow path. At runtime the Switch node selects an output terminal based on the evaluation of a set of conditions that you specify in the Switch node.
In a terminal flow (see Terminal flow) you can create flow branches using an Invoke screen operation node with multiple output terminals, each output terminal connected to another Invoke screen operation node. At runtime the node selects an output terminal based on which application screen is sent by the terminal application running on the remote host.
You can create loops in the following ways:
In either a nonterminal flow or a terminal flow you can create a loop using a While node (see Basic information about While nodes).
In a terminal flow, in addition to creating a loop by the method described in the previous bullet, you can also record a loop during flow recording (see Recording loops).
This section applies only to a terminal flow (a flow that interacts directly with a terminal application on a remote host).
In the host editor:
You can create a new terminal flow and record the flow path into the new terminal flow (see Recording a flow).
You can record one or more additional paths into an existing terminal flow that already contains a flow path (see Recording a flow).
You can play back the flow path of a recorded terminal flow (see Playing a terminal flow).
A path (or disconnected path) is a connected sequence of nodes that is not connected in any way to the flow path of a flow.
During the development and testing of a flow, it is common for a flow to temporarily contain one or more paths that the designer has not yet connected to the flow path. These disconnected paths are saved into the flow file when you close the flow editor and are displayed again on the flow editor canvas when you reopen the flow file in the flow editor.
However, when you run the Generate Runtime Code wizard on a flow containing such disconnected paths, the wizard does not process the disconnected paths and does not generate any runtime code for them. Therefore, as far as the runtime environment is concerned, disconnected paths do not exist at all in the flow.
Therefore, for a path to be executable at run time, it must be connected in some way to the flow path of the flow.