When it's your job to fix the 15,000 errors picked up after suggesting to add ESLint to an existing project 😣

You're making the code better? Why do you complain?

Cause it's alot to make better lol

Now you will think twice before suggesting a linter again haha!

Autofix?

This is a very common problem (large code base, thousands of findings).
At my company, we often recommend to use the so-called "boyscout rule", which means: Fix the warnings in the code you touch anyway". (This works best with incremental static analyis)

@crisz Yeah, that's 15,000 after --fix 😧

@losdanielos I've tried implementing that at a few places and found whether it works is based heavily on the company culture and the types of devs you're dealing with.

At the places where devs care about code quality and the Business understands the value spent improving code quality, it can work.

Unfortunately, some companies and devs have a "if it just about works then ship it asap".

You also have the issue of many devs being underqualified for their positions and not being able to do a simple refactor from an imperative to declarative loop or write pure functions or understand type coercion, etc...

@terminalterror totally agree that it heavily depends on the company culture. My company specializes in software quality and maintainability, and we try to practice what we preach in the development of our tool, so I'm lucky ;)

You could use something like Beautify to take care of a lot of formatting things, that might make things a little easier.