Fluid Architecture

流体架构

• 批准号: 9504339
• 负责人: William Scherlis
• 金额: $ 25万
• 依托单位: Carnegie-Mellon University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1995
• 资助国家: 美国
• 起止时间: 1995-07-01 至 1998-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9504339&HistoricalAwards=false
• 关键词: Fluid; Architecture

This project explores the design of systematic techniques to adapt the internal interface and component structure of programs of moderate size or larger. Programming tools based on these techniques could have the effect of increasing the extent to which the internal interface structure of systems can be made more fluid, that is, evolvable and adaptable. One potential application is to reorganizing abstraction boundaries in programs: It is clear in practice that software structural architectures are rarely optimal, but the costs and risks of adapting them are high. Another potential application is to the tailoring of general purpose frameworks and components to specific constrained uses, improving performance. Four questions are addressed: (1) What semantics-based program manipulation techniques are needed to support manipulation of internal software interfaces represented, for example, as encapsulation boundaries in programs? (2) How can soundness and other properties of the program manipulation techniques be established? (3) What are the design considerations for the development of tools that can operate on programs of increasing complexity and size? (4) How can the utility of the techniques be assessed for development and re-engineering of internal software interfaces and associated components?

这个项目探讨了系统技术的设计，以适应中等规模或更大的程序的内部接口和组件结构。基于这些技术的编程工具可以增加系统内部界面结构的流动性，即可进化性和适应性。一个潜在的应用是重新组织程序中的抽象边界：很明显，在实践中，软件结构架构很少是最佳的，但适应它们的成本和风险很高。另一个潜在的应用是将通用框架和组件定制为特定的受限用途，从而提高性能。 讨论了四个问题：（1）需要什么基于语义的程序操作技术来支持对内部软件接口的操作，例如，在程序中表示为封装边界？ (2)如何建立程序操作技术的可靠性和其他属性？ (3)对于开发能够在日益增加的复杂性和规模的程序上运行的工具，在设计上要考虑什么？ (4)如何评估这些技术在开发和重新设计内部软件接口和相关组件方面的效用？