Recombinant and evolutionary high-level software architectures: Design, integration, evolution and analysis

软件工程的基本目标之一是通过组装软件体系结构组件来实现软件系统的可预测和模块化构建和演化。然而，由于当前信息系统的日益复杂、分布式和动态的性质，来自内部违规的未检测到的集成故障最终可能仅被揭示为通常导致不期望的系统崩溃的代价高昂的事故。此外，接口和配置问题经常是由误解、不兼容性或由独立开发实践、异类环境或不同供应商引起的差异引起的。因此，由于维护和更改这类(基于Web的)分布式动态软件系统通常需要检查和修改组件以及组件交互，因此随着系统组件的扩展或升级，系统集成变得难以维护。这个项目的主要目标是定义和评估一种新的重组和演化软件体系结构方法，以及支持基于组件的设计、集成、演化和分析的方法和工具。该方法基于分布式基于事件的重组组件、扩展的组件接口、基于关注点和过程的演化扩展以及集成和演化契约的形式化抽象模型，可以提供基于组件的系统设计表示的更具表现力的定义，以及对体系结构组件的交互、其扩展的功能和行为及其组合的增强分析。总体而言，这项研究的结果可以影响正在努力生产大型、复杂(动态)系统和应用程序的个人或(加拿大)组织，并有助于推动提高生产率、软件质量、减少维护管理费用的努力，以及对建立基于合同的软件组件市场的影响。此外，我们相信，提供非特别和全面的组件集成和演化理论及其自动化支持将从根本上改进软件工程工作。

One of the fundamental goals of software engineering is to enable predictable and modular construction and evolution of software systems by assembling software architecture components. However, due to the increasing complexity, distributed, and dynamic nature of current information systems, undetected integration failures from internal violations may be ultimately revealed only as costly accidents that often cause undesirable system crashes. In addition, interface and configuration issues frequently arise from misunderstanding, incompatibilities, or differences resulting from independent development practices, heterogeneous environments, or different vendors. As a result, since maintaining and changing this class of (Web-based) distributed and dynamic software systems usually requires the examination and modification of components and component interactions, system integration becomes difficult to maintain as system components are extended or upgraded. The main goal of this project is to define and evaluate a new recombinant and evolutionary software architecture approach, and supporting methods and tools, for component-based design, integration, evolution, and analysis. The approach is based on formal abstract models of distributed event-based recombinant components, extended component interfaces, concern and process-based evolutionary extensions, and integration and evolution contracts that can provide a more expressive definition of component-based system design representations and an enhanced analysis of the interaction of architectural components, their extended functionality and behavior, and their compositions. In a general way, the results of this research can impact individuals or (Canadian) organizations that are struggling to produce large, complex (dynamic) systems and applications, and can help push the drive to improve productivity, software quality, reduce maintenance overheads, and impact on the establishment of a marketplace of contract-based software components. In addition, we are convinced that providing a non-ad-hoc and comprehensive component integration and evolution theory and its automated support will radically improve software engineering efforts.