Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Cost-benefit optimized quality assurance strategies for software product lines

: Kolb, R.

Volltext urn:nbn:de:0011-n-313810 (1.1 MByte PDF)
MD5 Fingerprint: bcae984cfbc923d1d73bfebc191cecf8
Erstellt am: 20.09.2005

Kaiserslautern, 2005, VIII, 84 S. : Ill., Lit.
IESE-Report, 059.05/E
Reportnr.: 059.05/E
Bericht, Elektronische Publikation
Fraunhofer IESE ()
quality assurance; software product line; testing; Café; ITEA

This report is concerned with planning and preparation of quality assurance for product lines in general and testing in particular. It addresses the problem of selecting an appropriate quality assurance or test strategy in the context of product line development by taking into account costs and benefits. The goal is to support organizations developing product lines in selecting or defining the quality assurance strategy that is optimized with respect to costs and benefits. To this end, the report provides some practical guidance for understanding and computing cost and benefits from various quality assurance strategies. First, it provides an overview of quality assurance for product lines and outlines basic quality assurance strategies. Then, it provides a qualitative discussion of the major cost elements and time and cost drivers of quality assurance strategies for product lines and discusses the relationship between costs and benefits. Furthermore, the report proposes a basic economic model, which provides the necessary background for understanding the impact of selecting a certain quality assurance strategy. Using the model, quality assurance strategies for product lines may be systematically measured or estimated and alternative strategies may be evaluated and compared for their efficiency and effectiveness. The report also presents a method for selecting the quality assurance strategy most appropriate for a given context and optimizing costs and benefits. Besides the model for costs and benefits and the method, the deliverable discusses the constraints on selecting a quality assurance strategy. Finally, it provides a discussion of the problem of having limited data in order to decide for a particular quality assurance strategy.