Flux, as far as I understood, is a technique about dealing with the dataflow of an application unidirectionally, keeping state isolated from the rest of the program in read-only, self-contained "stores" which can change its own state by listening to "actions" which are emitted by views and dispatched by dispatchers. Or, in short - a way to control state.
If that is right, how does it correlate with Functional Reactive Programming? Since FRP has a very strong control over state, I'd guess those are mutually exclusive techniques that actually solve the same problem. So, if one uses a FRP library (such as Elm), then there is little use to Flux. Is this correct?