A Methodology for Assessing Software Releasibility.
Abstract
A distinct challenge of Software engineering is the tradeoff between developing a high quality software product and delivering it on schedule. This thesis proposes a new methodology which addresses this tradeoff. The term "releasable" Software is introduced as a product which demonstrates a fault content acceptable to users in the field. The releasability assessment methodology capitalizes on basic testing metrics, software reliability modeling, statistical analysis techniques, and program specific criteria to present an objective estimation of the software release date. It is illustrate as an adaptive series of detailed procedures tailored to the unique needs and assumptions of the program. A division of Rome Laboratory recognized for medium-large scale software development provides the perspective for investigating finer points of the methodology. A notion of Configuration Reliability and the importance of system configuration management are presented. The impact configuration problems can have on software testing is discussed and a root cause analysis technique is recommended for achievement of optimal releasability.
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1997
- Accession Number
- ADA331874
Entities
People
- Matthew J. Kochan
Organizations
- Rome Laboratory