作者: Rodrigo RG Souza , Christina FG Chavez , Roberto A Bittencourt
DOI: 10.1186/S40411-015-0024-Z
关键词:
摘要: Writing patches to fix bugs or implement new features is an important software development task, as it contributes raise the quality of a system. Not all are accepted in first attempt, though. Patches can be rejected because problems found during code review, automated testing, manual testing. A high rejection rate, specially later lifecycle, may indicate with process. Our objective better understand relationship among different forms patch and characterize their frequency within project. This paper describes one step towards this objective, by presenting analysis large open source project, Firefox. In order rejection, we relied on issues commits from over four years project’s history. We computed monthly metrics occurrence three indicators rejection—negative reviews, commit backouts, bug reopening—and measured time takes both submit reject inappropriate patches. Firefox, 20 % contain Negative issue reopening relatively independent events; particular, about 70 reopenings premature; 75 changes days. Patch frequent event, occurring multiple times day. Given relative independence types, existing studies that focus single type fail detect many rejections. Although cause rework, they have little effect released versions