Using dataflow variables for lazy evaluation

Dataflow concurrency is a concurrent programming paradigm that has been around for three decades now. What is so exciting about it?

The main idea behind Dataflow concurrency is to reduce the number of variable assignments to one. A variable can only be assigned a value once in its lifetime, while the number of reads is unlimited. If a variable value is not written by a write operation, all the read operations are blocked until the variable is actually written (bind). With this straightforward, single-assignment approach, it is impossible to access an inconsistent value or experience data race conflicts. The deterministic nature of Dataflow concurrency ensures that it will always behave the same. You can run the same operation 5 or 10 million times the result will always be the same. Conversely, if an operation enters into a deadlock the first time, it will do the same every other time you run it. These qualities make it very easy to reason about...