How does quality deviate in stable releases by backporting?

ICSE (NIER)(2023)

引用 1|浏览16
暂无评分
摘要
Software goes through continuous evolution in its life cycle to sustain bugs and adopt enhanced features. However, many industrial users show reluctance to upgrade to the latest version, considering the stability and intuitive solace of the release they are using. This boosts the need to derive change patches from state-of-the-art versions to older software versions. This phenomenon is frequently supported by 'Backporting' in the industrial setting as the intent for backward patch propagation stood principally to sustain older releases, and the contribution does not count up to the upstream repository. However, it is yet unknown whether backport can act as a credible threat for stable releases. In this study, we aim to empirically quest backports to reveal the evolution trend of code entities through maintenance and pinpoint how they pull stable releases into the weak spectrum. The breakdown shows code entities often encounter gradual transformation in size, complexity and coupling due to consecutive commits on them. However, the numerics of outlier quality degradations are not insignificant at all in this context which calls for further investigation into why and when they may occur. Moreover, we observed that vulnerable change transmission often materializes with quality degradation. Understanding these issues and consequences is crucial for effectively supporting the backporting process for stable release maintenance.
更多
查看译文
关键词
Code Quality,Backporting,Software Engineering Software Maintenance
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要