Alternate Modification Measurement Method And Tool Implementation For Open Source Software Ecosystem

With the rapid development of software technology and people's growing enthusiasm for programming,community development which is open,flexible and efficient obtained the recognition and popularity of the world,a large number of open source software which pooled the wisdom of crowds emerged and are being widely used.In such development trend,many participants are mutually cooperating and their interests are interrelated,which gradually formed various open source software ecosystems.Due to the large scale of the open source software ecosystem,complicated participants' activities,the various development abilities of developers and different requirements that they are interested in,some compromised technical solutions are easily adopted,which will lead to the accumulation of technical debt of open source software.At the same time,because of collective code ownership,most of the source files of open source projects have been modified alternately by different developers irregularly,for later developers,they have to spend more effort to understand the code which had been modified by other developers,this will cause deviations in understanding,so as to introduce more quality issues.Existing research that investigated the impact of changes made by multiple developers on software quality in the level of the open source software project were mostly empirical works,which lack of in-depth analysis of developer activity relationship,and did not consider that different developers modify source code alternately will likely introduce more issues such as defects and technical debt.Meanwhile,technical debt as a new perspective to understand software quality and maintainability,the methods used to identify and measure technical debt are rather limited.Therefore,it needs an effective way to consider quantitatively the influence of alternate modifications made by different developers,to some extent,to indicates the level of the potential technical debt in open source software.Faced with the limitations of the above research,this paper first describes a measurement method used to evaluate the degree of alternate modifications of open source software in detail,then designs and implements an automatic analysis tool to calculate alternate modification index and defect density of source files,then a large-scale case study is designed followed by a set of rigorous experiment procedures.In order to avoid the uncertainty of experiment results,ten large-scale Apache open source projects of Apache Open Source Software ecosystem are chose as study objects,and making full use of submit records,source code and issue information provided by GitHub and Apache issue tracking system,to analyze and calculate Spearman correlation between alternate modification index and defect density of the source files using the tools developed and SPSS software.The study results show that alternate modification index of source file has a significant positive correlation with its defect density,which means that in the development process of open source software,if multiple developers alternately modify the same source file frequently,it will be easier to introduce more defects and technical debt in this source file,as a result,this will negatively affect open-source software's quality,maintainability and comprehensibility.