high-level and platform-independent description of digital signal processing (DSP) algorithms. Feldspar is a pure functional language embedded in Haskell. It offers a high-level dataflow style of programming, as well as a more mathematical style based on vector indices. The key to generating efficient code from such descriptions is a high-level optimization technique called vector fusion. Feldspar is based on a low-level, functional core language which has a relatively small semantic gap to machine-oriented languages like C. The core language serves as the interface to the back-end code generator, which produces C. For very small examples, the generated code performs comparably to hand-written C code when run on a DSP target. While initial results are promising, to achieve good performance on larger examples, issues related to memory access patterns and array copying will have to be addressed. I.

NASA is using model-based languages and risk analysis methodologies to raise software development to the level of hardware development. Ultimately, it hopes to achieve a fusion of systems and software engineering by replacing conventional software development techniques with capability engineering, which focuses on a system’s full set of functionalities. Many NASA flight systems have been developed according to an approach that defines hardware first, then effectively retrofits software and human procedures to the hardware systems. Throughout this process, a careful allocation of functionality handles time-sensitive operations onboard, while mission controllers on the ground handle non-time-sensitive operations. Past missions have been scripted, meaning that engineers developed hardware and software systems to respond mainly to foreseen circumstances, making them less able to handle unforeseen problems or exploration opportunities. The primary responsibility for handling unanticipated situations resided with humans, either onboard the spacecraft or in the mission control center. Clearly, there are times when even time-sensitive concerns are handled on the ground because of the more substantial human and computer resources available to address complex problems. All the software support for onboard and mission control must be highly dependable, requiring rigorous validation and verification. The US Department of Defense recently presented data showing that the percentage of aircraft functionality provided by software increased from 8 percent for the F-4 in 1960 to 80 percent for the F-22 in 2000. 1 Although software technologies supporting the development of onboard systems have improved significantly in the past 45 years, whether improvements have kept pace with the growing demand and complexity that result from the increasing software functionality remains arguable. Following current practices, it can take years or months to make software modifications to spacebased systems at significant cost. This remains true even though the personnel who develop and maintain