Figure 5 also shows that “suboptimal solution”, “duplication” and “incomplete fix” are in general decisive for rejecting a patch, while “violating coding styles” and “bad naming” are considered to be less decisive.
In addition to the patch-rejection reasons already included in the survey, 44 out of the 246 respondents supplemented additional reasons that they considered important. The topmost mentioned reason was “no accompanied test cases”, as 9 respondents considered a patch submitted without test cases to be unacceptable. “Performance” and “patch-writers’ expertise and reputation” were both mentioned by 5 respondents.