Behind the scheduled delivery of poor-quality products overrunning the cost or budget are the major blames the software development industry has to handle even today, at its apparent peak period of excellence at present. In spite of continuous refinement over the last four-five decades, it is still not free from many genuine criticisms like these. To get rid of these criticisms gradually, proper evaluation of different software development controls and measures, and timely elimination of code defects in each stage of data platform application development are required to ensure better management of software development life cycle and improved performance of the launched software. This article presents a simulation model of a data platform using code inspection and code scorecard and comparing the removal rate of code defects. Results of the simulation model prepared by the authors show a significant reduction in code defects, improved software development processes, and flawless production execution, which is expected to enable the software developers to improve their current processes.
TopIntroduction
In spite of continuous refinement over the last four-five decades, the software development industry is still not free from many genuine criticisms in terms of providing the best or perfect possible products & services up to the level of high expectations of the increasingly growing software/application users around the world. On the basis of the experiences of working on or being involved in different important software application development projects as an industry professional, and an academician respectively, the authors have been searching for suitable answers to these criticisms.
In the process, from the similar research, or related studies done by the authors in the recent past, they identified the major critical areas which need to be addressed relentlessly, and more & more thoroughly always. Those are i. Controlling Cost, ii. Maintaining Schedule, iii. Ensuring Quality (Khanam and Rahman, 2019). There are many tits and bits to deal with to address these major issues. The authors attempted to develop a simulation model showing the possible solutions to the root problems in each of these major areas by investigating the resolved and unresolved challenges faced by the software development projects undertaken by the renowned supplier of software application programs in the world. The prepared simulation model for application development in Enterprise Data Platforms (EDPs) aims to analyze all the related & root causes of criticism to ensure an efficient life cycle of software development, optimized cost throughout the process, and improved performance of the launched & marketed applications.
It is important that code defects at each stage of software development need to be scrutinized very strictly. This paper illustrates it by adding two extra phases viz. code inspection and code score carding to the usual phases of software application development life cycle in the Enterprise Data Platforms (EDPs). With the proven experiences gained from other similar studies done by the authors in the past, it can be claimed that defect removal rate is much higher in the process having these two extra phases compared to the usual process not quite caring about code inspection and code score carding in every major stage of software development. It is highly expected that the simulation model presented in this paper would certainly enable the developers to improve their existing practices in giving quality and timely output of application software in EDPs (Dabab et al., 2018).
The authors have revised, enhanced, and updated their related research papers for the purpose of preparing this paper to present the said model. The due effort was given to bring appropriate answers to the following questions in the simulation process:
-
1.
How much improvement may be made in process performance by introducing source code inspections?
-
2.
Does the error detection capability of source code inspection leave any impact on the software development process in EDPs?
-
3.
What are the risks of not going beyond the lowest rate of error detection even after adding the phase of code inspections?
-
4.
How to choose only one between source code inspection and unit test?
-
5.
How to justify the benefit of source code inspections if the starting source code shows more errors?
-
6.
How to compare the existing cost of any incorrect unit test done in a software development process in EDPs?