Technical Paper, ROCAS v1.2 c 1060 Research Ltd Abstract Today's software systems have a much shorter life span than other engineering arti- facts. Code is brittle and system architectures are rigid and not robust to change. With the Web an alternative model of computing has taken shape, a model that allows for change on a small scale inside of a complex and large-scale system. The core abstrac- tion behind the Web can be applied to the design of software systems. The abstraction in Resource Oriented Computing generalizes the Web abstraction to enable the most adaptable software systems and architectures that evolve with change. 1 Resource Oriented Computing for Adaptive Systems Peter J. Rodgers, PhD,
[email protected] May 2015 1 Challenge The second Law of Thermodynamics states that entropy will always increase. The challenge faced by long-lived resource adaptive systems is to fight entropy with the least possible expenditure of energy (physical, computational, financial). 2 Definitions "Myriad computational systems, both hardware and It is helpful to start by stating some definitions so that we may clarify concepts such as software, are organized as \stability" and \evolvability" as applied to adaptive computation systems. state-transition systems. Such a system evolves over It will be shown that we can obtain a useful working understanding of stability and time (or, computes) by evolvability by considering computational state.1 continually changing state We require to maintain a conceptual distinction between a given logical abstract potential in response to one or more discrete stimuli (typically state and an instantaneous physical representation of the state.