Published on Oct 11, 2013
From the new book, "Dataflow and Reactive Programming Systems"
"Dataflow and Reactive Programming Systems" - A Practical Handbook on Developing Dataflow-Like Programming Systems
Dataflow concepts are the heart of Reactive Programming, Flow-Based Programming (e.g. NoFlo), Unix pipes, Actors and message passing in general.
Dataflow-based systems are easy to design once you understand the large number of implementation details that could drastically change how the system operates. Understanding these vectors of change is important so you don't waste your time developing the wrong system.
Embedded dataflow-like languages are used in a wide range of applications. Video games, web pages, circuit simulation and music production are just a few of the domains that have been using dataflow for years. Every one of those has a specialized dataflow engine designed for the task at hand. This book will help you understand the whole dataflow universe before starting your own system.
By the end of the book you will understand...
- All possible design choices with dataflow-like systems
- How their effects interplay
- How to develop your own dataflow-like system
Public domain source code of an example dataflow system will be available for download from the book's web site once completed.
Table of Contents (Preliminary)
General Architecture: (Brief introduction to dataflow and its terminology)
- Benifits of Dataflow
- Dataflow in a Nutshell
The Nouns of Dataflow: (Definitions of the common parts of dataflow-based systems and their possible design choices)
- Nodes & Ports
The Verbs of Dataflow: (Covers the design choices for the runtime operation of dataflow-based systems and their affect on its operation)
- Scheduling & Executing
- Time (Glitches & Synchronization)
Problematic Graphs: (Understand how implementation choices can change the operation of certain graphs -- programs)
Specific Implementations: (Describes the key characteristics of common dataflow models with sample code for each section. Discusses how the design choices affects operation -- its pros and cons in comparison to other dataflow models)
- Pure Dataflow Programming
- Flow-Based Programming
- Reactive Programming
- Unix Pipes
- Service Oriented Architecture
Code samples in C# and other major languages:
- Download code of a complete dataflow-based programming system that you can extend
- The .Net Reactive Extensions will not be used
- Instead, we will show you an example of a complete system built from first principles