The Characteristics Of Software Quality

Published Date: 02 Nov 2017

Software Quality is the entirety of description and characteristics of an artefact or service that stand on its capability to assure positive or implied requirements (Akingbehin 2005). Quality can establish a software product’s achievement or failure in competitive economic market.

There are two major sites when considering the meaning and definition of software quality (Hoyer 2003):

Conformance to specification: Quality that is classified as a subject of products and services whose evaluation features satisfy a fixed specification.

Meeting customer needs: Quality that is defined independent of any assessable features. That is, quality is identified as the products or services potential to satisfy customer expectations.

Characteristics of Software Quality (Pressman 2010)

According to Pressman, Hewlett-Packard came up with a set of attributes of software results by which its excellence is expressed and evaluated, they have been given the acronym FURPS;

Functionality – is measured by evaluating specific properties that satisfy stated needs.

Usability – is measured by effort needed for use (learnability), consistency and overall aesthetics. A system does not have usability features if the system is not functioning perfectly

Reliability – is assessed by the measure of the rate of recurrence and severity of breakdown, correctness of product, the mean-time-to-failure, the capability of picking up from crash (recoverability), and the predictability of the system.

Performance – is evaluated by considering processing speed, respond time the quantity of resources used, throughput and effectiveness.

Supportability – involves extensibility, adaptability, serviceability, testability, compatibility and configurability.

Causes of Software Poor Quality

Lack of domain knowledge:

Lack of domain knowledge is a contributing fact to poor software quality; the majority of developers are not specialists in the business area served by their applications, e.g. telecommunication area, banking, supply chain and retail.

To moderate this cause is that experts from the business, should train developers in the business areas, and carry out peer reviews with individuals who have domain more knowledge.

Lack of enough technology skills:

The majority of developers are expert in quite a few programming languages. Current multi-tier commerce applications cover many computer languages and various software platforms.

This can be mitigated by cross-training developers in various application technologies, carry out peer reviews with developers from other application tiers and carry out static and dynamic study of the system.

Impracticable schedules:

If sound software development exercises are forced with unreasonable schedules, developers end up producing bad results.

When operating at fast pace, pressured developers make more mistakes and have no time to locate them. A means to mitigate these is by imposing strong project management practices; Managing commitments by planning, tracking progress to identify issues, and managing continual requirements changes.

Badly engineered software:

Most software improvement action involves varying or enhancing existing code. Half of the duration exhausted modifying existing software is spent with trial to understand the code.

Unreasonably complex code is often unsolvable and modifying it ends with frequent mistakes and unforeseen negative side effects. The recently introduced defects cause costly re-work and delayed releases. The most excellent way of mitigating this cause is to re-factor vital segments of the code directed by items for architecture and studying static system.

Inadequate acquisition exercises:

Well-built multi-tier functions are developed and taken care by distributed teams, some of whom might be outsourced from other companies. As a result, the affected organization often has less visibility into or managing the quality of the system they will be using.

CMMI levels have not always assured high quality software release. To mitigate the risks of quality issues in externally supplied software, managers should apply quality objectives within the agreements and a definite quality guarantee gate for delivered software.

SOFTWARE MEASUREMENT

Upgrades in technology are continually increasing the demand of software that is well-built, more robust and more consistent over ever-widening series of application. The demands on software management are rising correspondingly. Software developers, managers and technical staff are frequently confronted with new technologies and its challenges, more competitive markets, increased competition for knowledgeable personnel and demands for faster receptiveness. They carry on to be worried about open-ended requirements, unmanaged changes, inadequate testing, insufficient training, subjective schedules, insufficient funds and standards related issues, product consistency, and product suitability.

Software measurement alone cannot work out these problems, but it can make clear and focus your understanding of them. When done appropriately, sequential measurements of quality features of products and processes can offer an efficient foundation for initiating and controlling process development activities. Measurement is the practice by which numbers and symbols are established to characteristics of objects in the real world, this permits us to identify such objects according to identified rules. With software development, measurements are carrying out by using metrics; metrics are experimental description of a value to an object trying to set a specific quality of this object.

Benefits of Software Measurement

Locates opportunities for upgrading

Improves communication

Encourages proper behaviour

Communicates workloads

Leverage resources

Enhance profit on IT investments

Decrease cost

Control workloads

Decrease overtime

Boosts requirements process