Reactive Information Processing

People agree that there are two major concerns in Internet computing: Big Data and the Internet of Things (IoT). On purpose, expected evolutions and progresses in technology and science are ruled by the development of suited paradigms (e.g., plug & play middleware for the IoT or MapReduce for Big Data) to face up this “ever encountered” nature of Internet computing. To that extent, information processing (from raw data to meaningful - i.e., semantically rich-information) encompasses the necessity of building a considerable pool of Internet software in a truly different way. Such a paradigm shift is exposed in “The Reactive Manifesto” (www.reactivemanifesto.org). From an architectural perspective, this manifesto promotes software applications' componentization along with the idea of reactiveness: event-driven/message-driven, elasticity, responsiveness and resilience. In short, applications' components have emerging (reactive) features through their ability to seamlessly cooperate with events/messages. Nowadays, successes like Node.js or WebSockets strongly confirm the benefit of reactiveness. Beyond, this keynote tries to demystify and illustrate “reactiveness” through the State Chart XML W3 standard. The keynote discusses methods to design reactive Internet software from models to concrete implementation supports.