Software infrastructure is becoming an ever more important and costly enabler of scientific discovery. The investigators will gather and present input from the science, software developers, and software integrator communities concerning the functions and services (e.g., quality assurance, hosting, training, impacts assessment) that the NSF can provide that would improve the overall quality of the software infrastructure available to our scientists, reduce the cost-of-ownership of this software and maximize the return on the NSF investment in developing, maintaining and evolving this infrastructure.
The three communities will be engaged through back-to-back focused workshops that will bring together selected experts from a broad range of fields and disciplines to discuss the life-cycle of software infrastructure. The workshops will generate reports that input to and give requirements for the upcoming NSF's call for Scientific Software Institutes. The first workshop will address the software requirements and expectations of domain scientists whose computational infrastructure consists of distributed resources, including computers, data, and instrumentations as illustrated, for example, by projects such as the LHC experiments, OSG, LIGO, Earth System Grid, SCEC, NEES, caBIG, and BIRN. The collective experiences and drivers from software integrators who have been providing and supporting the software infrastructure for such projects will be discussed in the second workshop. The third workshop will cover the needs and expectations of projects that develop and maintain widely adopted scientific software tools. The three dimensions of scientific software infrastructure will be covered throughout the workshops' vertical that addresses end-to-end requirements, horizontal that accommodates reusability and broad impact, and the software providers who are responsible for developing, supporting and sustaining the software artifacts.
Project Outcome Report from the Workshops on Distributed Computing, Multidisciplinary Science, and the NSF’s Scientific Software Innovation Institutes Program Miron Livny, Ian Foster, Ruth Pordes, Scott Koranda, JP Navarro October 2010 The workshop resulted in recommendations to the NSF based on the input from a broad set of experts in the field. These recommendations have further been stimulating the thinking and directions of the participants in the workshop. The recommendations included the main outcome below as well as a set of supporting more specific recommendations published in the final report.[1] Outcome – Recommendation 1 The most significant outcome of the workshops was the vision (and key attributes) for: "A US Software Infrastructure Institute that provides a national center of excellence for community based software architecture, design and production; expertise and services in support of software life cycle practices; marketing, documentation and networking services; and transformative workforce development activities." The measure of success of such an institute should be the cost effectiveness (as measured in scholarly work) of our software infrastructure. This will be accomplished by a thriving and innovative software infrastructure community anchored by the Software Infrastructure Institute. The Institute will play a unique role by addressing organizational and life cycle elements not covered by infrastructure implementation and deployment projects that are driven by scientific objectives or technological trends. [1] http://osg-docdb.opensciencegrid.org/0010/001002/001/WorkshopReportSubmittedtoNSF.pdf
