2 ms·

This is a click-bait headline coupled with a terrible analysis. I don't even know what "bug density" is because the author doesn't define it, but I assume it i

This is a click-bait headline coupled with a terrible analysis. I don't even know what "bug density" is because the author doesn't define it, but I assume it is some count of bugs divided by lines of code. It doesn't seem like the author tried to normalize or control for any factor, including basic ones like the age of the project. I'm not sure what a sound approach to answering this question is—I think this is a topic worthy of academic study—but this is definitely a low-quality analysis.

yawaramin 9y ago
The author defined it after responding to comments--he says it's bugs per repo. So, a repo with 10 lines of code and 1 bug is, according to him, less buggy than a repo with 100 lines of code and 2 bugs. Also, note that projects in dynamically-typed languages tend to be smaller since people tend to switch over to static types for larger projects. So this 'analysis' is pretty much designed to 'find' that dynamically-typed language projects are less buggy.