There's no elasticity over what the OP reported. That was a bug. Clear as they come. He may have unit tests but they didn't catch it and since you can't have unit tests for everything, why not spend 500ms running a linter?
Testing and linting does different things. Of course a sandwich is cheaper than a condo, but what does that have to do with what we are talking about?
As for the claim that you can't have unit tests for everything: What do you mean? No, you can't test all branches or combinations in a system, but you can have a 100% test coverage (aka. hitting each line at least once).
Yeah not really. Both linting and unit tests catch bugs. This is a fact that you refuse to actually face.
>you can have a 100% test coverage
You can but it's A) a bad idea and B) it won't even necessarily catch the bug that the OP brought up if you do.
I use both linting (pylint actually) and unit tests, so I am not advocating that you should scrap your linting tools. I am just trying to point out the difference.
Your B) argument is yet another indication that the tests are not good enough. The example bug OP posted is what made me comment in the first place. If your tests don't catch that your tests are useless.