Bibliographic Citation
See/Add Document Discussions Return to Search Results Return to Original Search Page Download as EndNote
Full Text	936 K   View Full Text or Access Individual Pages  -   search, view and/or download individual pages
DOI	10.2172/15150
Title	Towards an Approach to Overcome Software Brittleness
Creator/Author	OSBOURN,GORDON C.
Publication Date	1999 Nov 01
OSTI Identifier	OSTI ID: 15150
Report Number(s)	SAND99-2850
DOE Contract Number	AC04-94AL85000
DOI	10.2172/15150
Other Number(s)	TRN: US200221%%249
Resource Type	Technical Report
Resource Relation	Other Information: PBD: 1 Nov 1999
Coverage	Topical
Research Org	Sandia National Labs., Albuquerque, NM (US); Sandia National Labs., Livermore, CA (US)
Sponsoring Org	US Department of Energy (US)
Subject	99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; COMPUTER CODES; DESIGN; MODIFICATIONS; PROGRAMMING; PROGRAMMING LANGUAGES; RELIABILITY; SAFETY; SECURITY
Description/Abstract	Development of bug-free, high-surety, complex software is quite difficult using current tools. The brittle nature of the programming constructs in popular languages such as C/C++ is one root cause. Brittle commands force the designer to rigidly specify the minutiae of tasks, e.g. using ''for(index=0;index>total;index++)'', rather than specifying the goals or intent of the tasks, e.g. ''ensure that all relevant data elements have been examined''. Specification of task minutiae makes code hard to comprehend, which in turn encourages design errors/limitations and makes future modifications quite difficult. This report describes an LDRD project to seed the development of a surety computer language, for stand-alone computing environments, to be implemented using the swarm intelligence of autonomous agents. The long term vision of this project was to develop a language with the following surety capabilities: (1) Reliability -- Autonomous agents can appropriate y decide when to act and when a task is complete, provide a natural means for avoiding brittle task specifications, and can overcome many hardware glitches. (2) Safety, security -- Watchdog safety and security agents can monitor other agents to prevent unauthorized or dangerous actions. (3) An immune system -- The small chunks of agent code can have an encryption scheme to enable detection and elimination of unauthorized and corrupted agents. This report describes the progress achieved during this small 9 month project and describes lessons learned.
Country of Publication	United States
Language	English
Format	Size: 11 pages
Availability	OSTI as DE00015150 To purchase this media from NTIS, click here
System Entry Date	2007 Apr 23
Document Discussions Discuss	User Name:  (for display) Email Address:  (Email address will NOT be displayed.) Discussion Subject:    (All fields required. Document Discussions not displayed until approved.)
See/Add Document Discussions Return to Search Results Return to Original Search Page Download as EndNote
Top