#30 Dataflow model guide is ported from cpp.react (WIP)

Author: YarikTH

doc/dataflow_model.md

@@ -0,0 +1,80 @@
+# Dataflow model
+
+Disclaimer: based on [Dataflow model guide from cpp.react](http://snakster.github.io/cpp.react/guides/Dataflow-model.html)
+
+* [Graph model](#graph-model)
+  * [Input and output nodes](#input-and-output-nodes)
+  * [Static and dynamic nodes](#static-and-dynamic-nodes)
+  * [Domains](#domains)
+  * [Cycles](#cycles)
+* [Propagation model](#propagation-model)
+  * [Transactions](#transactions)
+  * [Properties](#properties)
+* [Conclusions](#conclusions)
+
+## Graph model
+
+The dataflow between reactive values can be modeled (and visualized) as a [directed acyclic graph (DAG)](https://en.wikipedia.org/wiki/Directed_acyclic_graph).
+Such a graph can be constructed from the dependency relations; each entity is a node and directed edges denote data propagation paths.
+
+To give an example, let `a`, `b` and `c` be arbitrary signals.
+`x` is another signal that is calculated based on the former.
+Instead of invoking `make_signal` explicitly, the overloaded `+` operator is used to achieve the same result.
+
+```cpp
+ureact::signal<S> x = (a + b) + c;
+```
+
+This is the matching dataflow graph: 
+
+<p align="left"><img src="../support/data/signals_1.svg"></p>
+
+From a dataflow perspective, what kind of data is propagated and what exactly happens to it in each node is not relevant.
+
+µReact does not expose the graph data structures directly to the user; instead, they are wrapped by lightweight proxies.
+Such a proxy is essentially a shared pointer to the heap-allocated node.
+Examples of proxy types are `signal`, `observer`.
+The concrete type of the node is hidden behind the proxy.
+
+We show this scheme for the previous example:
+
+```cpp
+ureact::signal<S> a = make_var(...);
+ureact::signal<S> b = make_var(...);
+ureact::signal<S> c = make_var(...);
+ureact::signal<S> x = (a + b) + c;
+```
+
+The `make_var` function allocates the respective node and links it to the returned proxy.
+Not all nodes in the graph are bound to a proxy; the temporary sub-expression `a + b` results in a node as well.
+If a new node is created, it takes shared ownership of its dependencies, because it needs them to calculate its own value.
+This prevents the `a + b` node from disappearing.
+
+The resulting reference graph is similar to the dataflow graph, but with reverse edges (and as such, a DAG as well):
+
+<p align="left"><img src="../support/data/signals_2.svg"></p>
+
+The number inside each node denotes its reference count. On the left are the proxy instances exposed by the API.
+Assuming the proxies for `a`, `b` and `c` would go out of scope, but `x` remains, the reference count of all nodes is still 1, until `x` disappears as well.
+Once that happens, the graph is deconstructed from the bottom up.
+
+
+### Input and output nodes
+
+### Static and dynamic nodes
+
+### Domains
+
+### Cycles
+
+## Propagation model
+
+### Transactions
+
+### Properties
+
+## Conclusions
+
+---------------
+
+[Home](readme.md#reference)

doc/readme.md

@@ -15,5 +15,6 @@ Usage:
 * [Tutorial](tutorial.md) - make sure you have read it before the other parts of the documentation
 -->
 * [Signals vs callbacks](signals_vs_callbacks.md)
+* [Dataflow model](dataflow_model.md)
 * [Integration](../tests/integration/) - how to get and link this library
 * [Examples](../tests/src/examples)