Assessing Growth and Instability of Open Source Software Systems

Software systems progressively evolve over time where source code needs to be modified, newly are added and others need to be deleted. Therefore, it is worth to monitor progress and software development, to measure growth and assess system stability. This paper assesses OSS software growth and instability to clarify impacts of its instability on its system growth. Three open source software applications, with different characteristics, have been selected. Size and instability metrics are extracted and statistically analyzed in order to assess the reflections of software growth under instability behaviors.The correlation among size metrics are highly correlated - the highest and lowest correlation are 0.99 and 0.72 respectively. Moreover, the result shows the highest correlation is between total classes and total lines of code. The instability measures on software systems exhibits certain pattern characterized by more addition and modification than deletion at system design level and the implementation changes exhibits more addition than modification and deletion. although the characteristics of the open source applications are different than the closed-source, the Lehman's observations are still valid with regard to continuous growth and trends.