An open source tool is being readied for release early next year that its creators say could dramatically speed software development and improve software quality. Developers this month are being invited to test a beta version.
Computer scientists from the National Institute of Standards and Technology (NIST) and the University of Texas at Arlington credit the use of "combinatorial testing" for their breakthrough.
The trick is being able to quickly test interactions of up to six variables. The work stemmed from research into what really causes bugs in software. The researchers found that it is more often due to problematic interactions between a few variables rather than a bunch even if a program, such as an e-commerce application, features hundreds of variables.
Combinatorial testing is used in other disciplines, such as chemistry. A researcher last year even got a £100,000 National Science Foundation grant to study Rubik's cubes as a way to learn about combinatorial problems in science and operations.
Findings of this latest software debugging research are described in several presentations, one by NIST researchers and another by University of Texas researchers.
Developers interested in getting their hands on the beta code should contact NIST's Raghu Kacker at [email protected] After the beta phase, an executable version will be released, followed by the open source code.