Multi Release Cost Model — A New Perspective
Gaurav Mishra, P. K. Kapur, Avinash K. Shrivastava · International Journal of Reliability Quality and Safety Engineering · 2017
Software testing plays a crucial role in software development process to achieve high quality software. Increased demand for new features, market competition, time and resource limitation has shortened the software lifecycle. To be in the market, software companies have to release new versions of software with new functionalities. Many ideologies and cost models have been formulated in the past decades with the idea of when to release the software for single and multi-versions. In the existing literature, the total debugging cost for the first version and next versions includes the cost of debugging in the operational phase along with the testing cost. It is assumed that the software is supported until the operational phase is over. Also, it is considered that some of the remaining faults of the previous release are reported and removed during the testing period of the new versions, which shows the repetition of cost for debugging the remaining number of faults of the previous release and results as increased cost of the software for the next release. In the present paper, we provide a new perspective on the release cost. To overcome the above problem in this paper, we describe a generalized cost model in which the remaining faults of the current version are removed during the operational phase of the current version and partly in the testing period of the next version. Application of the proposed cost model is shown on a real fault counting data for multi-versions of a software.