This paper proposes a way to realize the idea of calm computing by adding a dynamic task model into the pervasive computing environment. This task model contains information about the actions to undertake to help a user realize his daily tasks. The task model’s mapping onto a deployment plan guides an internal adaptation mechanism, which helps applications to evolve without causing user distraction. In addition, a foraging technique (relocation) is proposed that allows for expanding an application’s computing space automatically whenever possible. This technique involves external adaptation mechanisms. Both adaptation mechanisms are driven by resource information and resource contracts that are negotiated between the middleware and the application components. This allows the middleware to do the adaptations automatically, realizing the idea of calm computing.

As computational complexity of systems continues to increase, the amount of maintenance required to keep them operational will also increase. Autonomic systems have the aim of reducing the amount of maintenance required by performing certain levels of maintenance themselves. This paper outlines the case of using dynamic AOP to implement such a system. The benefits and issues arising from using dynamic AOP will be looked at and discussed. 1

Abstract. Dynamic AOP has been identified as a useful technique to implement an auto-adaptive framework. To supplement this policies based upon Event-Condition-Action rules are used to specify when adaptations should be applied to the underlying system. However, for use in auto-adaptive systems it is advantageous if these policies allow certain relationship information to be specified to ensure the system never encounters undesirable interactions and adaptations are applied in a correct manner. This paper highlights the need for these relationships by giving a series of examples followed by a proposed solution to these problems.

This paper reports the development of the Concerto platform, which is dedicated to supporting the deployment of resource-aware parallel Java components on heterogeneous distributed platforms, such as pools of workstations in labs or offices. Our work aims at proposing a basic model of a parallel Java component, together with mechanisms and tools for managing the deployment of such a component on a distributed platform. Moreover, we strive to provide components with means to perceive their runtime environment, so they can for example dynamically adapt themselves to changes occurring in this environment. The Concerto platform was designed in order to allow the deployment of parallel components on a distributed platform. It additionally defines and implements an open and extensible framework for distributed resource discovery and monitoring in such an execution environment.

This paper presents Load and Let Link – a framework for flexible runtime loading and linking of procedural native code components. LLL has several novel aspects. First, it provides componentization without requiring an object-oriented language. Second, LLL performs linking at runtime, providing arbitrary code expansion, contraction and substitution. This enables (a) adaptive applications that can rewire themselves in response to dynamic conditions, (b) code patching for mission critical systems and (c) automatic overlaying in memory constrained environments. LLL is language neutral and orthogonal to current software development methodologies, thus providing the substrate for the next generation of develop once deploy anywhere software. In this paper, we present the LLL framework, its implementation on 32 bit x86 architectures and two case studies that showcase the capabilities of the framework.