I came across this blog post from Python/Django developer Matt Layman, where he compares 2 different cases of code reviews: one where people sat in a room and spent a lot of time talking about style issues, and one where style checking was automated on GitHub and people started talking about the actual changes and problems the code tries to solve. I recommend reading it.

One thing that really interests me is how did the style guide they were not able to automate look like. Maybe AI based style checkers like FYT could have worked?

2 side takeaways: isort and flake8 for Python, which I will try out soon as I’m using Python/Django myself and especially imports are a topic I need to look into and do some refactoring on, it easily can get messy.

I’m curious how the decision finding and agreeing on rules like flake8 look like in a team, it must be a process of its own.