Composing next generation software component systems
Each entity currently found on the web, such as XML data and software has the potential to become an active component. Active components are expected to replace documents, desktops, browsers, 3-tier architectures and complete applications within the next years. For this reason, suitable technology for uniform and easy component composition is required that would increase reuse of components and hence productivity for both end-users and software engineers. Urged by this, the EASYCOMP project developed robust composition techniques for active components. Based on aspect separation and composition, these new techniques are expected to allow users to powerfully adapt components using suitable meta-modelling and code generation methods. Additionally, the developed techniques will also offer dynamic composition of active components for adapting systems. One of the key issues involved was the lack of homogeneity of components in the different development stages that are construction time, assembly time, and runtime. This diversity in the availability of means for composition causes increased complexity of component-based software development. To tackle this, researchers developed a uniform composition and adaptation methodology that integrates three different composition models applicable over all stages of the development lifecycle of a component into one model. The individual models offer increased capabilities of reuse and exploitation of artefacts that have been constructed in later phases of the component lifecycle. The integrated model provides a common environment for the components development for each of the three models. Taking into consideration existing Commercial Off-The-Shelf (COTS) components, the meta-component methodology establishes a sound framework for a generic API to access component models developed by different companies. Thus, users are expected to have more options when combining different model components.
