Experimentation Platform for Detecting and Resolving Interoperability Problems

9708643本研究的目标是在软件体系结构级别设计并执行组件集成实验，以确定(1)存在哪些组件互操作性问题，(2)这些问题的解决方案是什么，以及(3)哪些集成策略可以执行这些解决方案。这项研究的价值在于它可以为开发人员提供软件组件之间潜在互操作性问题的早期知识。有了这些知识，开发人员就可以将集成视为一个设计决策，而不仅仅是一个实现任务。随着COTS应用程序使用的增加，开发人员可以在功能相似但与内部组件的互操作性问题较少的应用程序之间进行选择，从而加快了生产力。通过对软件体系结构、设计模式和集成实际应用程序的报告的研究，建议的研究将定义导致互操作性问题的软件体系结构特征的目录。集成策略将与它们各自的解决方案相关联，这些解决方案是由交错的组件特征引起的。基于知识的系统将作为一个实验测试平台，以确定基于预定义的特征间冲突解决方案的最佳集成策略。研究结果，即特征、冲突、解决方案和策略，将提供给开发人员，以增强设计决策和制定实施方案。＊＊＊

9708643 The goal of this research is to devise and perform component integration experiments at the software architectural level that determine (1) what component interoperability problems exist, (2) what are solutions to those problems, and (3) which strategies for integration can perform those solutions. The value of this research is that it can provide developers with early knowledge of potential interoperability problems among software components. With this knowledge, developers can treat integration as a design decision, instead of only an implementation task. With the increased use of COTS applications, developers can choose between applications with similar functionality but fewer interoperability problems with in-house components, speeding productivity. Using research on software architecture, design patterns, and reports on integrating actual applications, the proposed research will define a catalog of architectural characteristics of software that contribute to interoperability problems. Integration strategies will be associated with their respective solutions to the conflicts that result from interleaving component characteristics. A knowledge based system will serve as an experimental testbed to determine the best integration strategy based on predefined conflict resolutions among characteristics. The research results, i.e., characteristics, conflicts, solutions, and strategies, will be provided to the developer to enhance design decisions and to formulate implementations. ***