Improvements for concurrent call-by-need functional programming languages.

点击翻译按钮获取中文摘要

The top goal of the project are validations of K-improvement properties of transformations in concurrent call-by-need functional programming languages w.r.t. different cost functions (for example time and space). To reach this goal, we have to develop theoretical and practical tools to recognize K-improvements and to verify the corresponding properties. A special challenge is concurrency, since the development of the foundation and tools for call-by-need calculiis in a very early stage. A special feature is the inherent sharing in call-by-need, which automatically permits redundancy free evaluations, but at the same time makes high demands on correctness and optimization-proofsThe chances of the project are: 1. intuition- and experience-based K-optimizations can be really shown as K-optimizations for various cost functions. 2. K-optimizations in the deterministic scenario could be proved to be optimizing also under concurrency. 3. Novel and theoretically founded optimization variants, also w.r.t. combinations of time and space and perhaps other costs can be of great help to optimally utilize the computing resource and could also be labeled with a guarantee.