Computers and software systems are becoming ubiquitous in modern society. Worldwide users rely on individual and interconnected computers, as well as the global information infrastructure, such as the Internet and the World Wide Web (WWW), to fulfill their needs for information processing, storage, search, and retrieval. All these needs are met with the support of the underlying software. This reliance requires the software to function correctly over a long time, to be easy to use, and so on. In general, such requirements for high quality need to be satisfied by the people involved in the development and support of these software systems through various quality assurance activities, and the claims for high quality need to be supported by evidence based on concrete measurements and analyses. In general, people’s quality expectations for software systems they use and rely upon are two-fold:
- The software systems must do what they are supposed to do. In other words, they must
do the right things.
It requires that software to perform the right functions according to the specification. For example, image editing software is supposed to supply photo retouching capabilities, while not intended to calculate expenses. - They must perform these specific tasks correctly or satisfactorily. In other words, they must
do the things right.
This expectation basically requires that the software systems perform their intended functions without problems. In image editing software example, the system should supply the user abilities of removing lens flare or red-eyes from the photograph, instead of wiping out part of the image, taking too long to perform the operation, or refusing to perform it without proper justification. The focus of the related activities is to verify that the implemented software functions operate as specified.
The tasks for software QA and quality engineering are to ensure software quality through the related validation and verification activities. These activities need to be carried out by the people and organizations responsible for developing and supporting these software systems in an overall quality engineering process.
So, what’s the difference between software testing and quality assurance?
Many software systems nowadays are highly complex and contain millions of lines of source code. Examples of such large software systems can be found in virtually every product segment or every application domain, from various operating systems, such as commonly used versions of the Microsoft Windows and UNIX-flavor systems, to commercial software products, such as database products, to aviation and in-flight entertainment software used on Boeing 777, to defense related software systems, such as various command/communication/control (CCC) systems. Such large and complex systems typically involve hundreds or even thousands of people in their development over months or even years, and the systems are often to be operated under diverse, and sometimes unanticipated, application environments. Accompanying the size and complexity problems are the many chances for other problems to be introduced into the software systems. Therefore, dealing with problems that may impact customers and users negatively and trying to manage and improve software quality are a fact of life for people involved in the development, management, marketing, and operational support of most modern software systems.
The above factors make it virtually impossible or practically infeasible to achieve the complete prevention or elimination of software problems and related negative impact. Consequently, various software QA activities are carried out to prevent or eliminate certain classes of problems that lead to such negative impact, or to reduce the likelihood or severity of such negative impact when it is unavoidable. Software testing, in fact, plays a central, but not only, role among the software QA activities. By running the software system or executing its prescribed functions, testers can determine if the observed system behavior conforms to its specifications or requirements. If discrepancies exist between the two, follow-up actions can be carried out to locate and remove the related problems in software code, which may also include modifying the software design. Therefore, the detection and removal of defects through testing help reduce the number of defects in delivered software products, thus helping to achieve the quality goals. Even if no discrepancy is observed, the specific instances can be accumulated as evidence to demonstrate that the software performs as specified. Consequently, testing is the most frequently used means to assure and to demonstrate software quality.
Nevertheless, beyond testing, there are many other QA alternatives supported by related techniques and activities, such as inspection, formal verification, defect prevention, and fault tolerance. Inspection is a critical examination of software code or other artifacts by human inspectors to identify and remove problems directly, without resorting to execution. Fault tolerance prevents global system failures even if local problems exist, through various redundancies strategically designed and implemented into the software systems. Other QA techniques employ specific means to assure software quality.
In addition, all these QA activities need to be managed in an engineering process we call the software quality engineering process, with quality goals set early in the product development, and strategies for QA selected, carried out, and monitored to achieve these preset quality goals. As part of this overall process, data collected during the QA activities, as well as from the overall development activities, can be analyzed to provide feedback to the software development process for decision making, project management, and quantifiable quality improvement.