Git Commit Message

[mtk@lemmy.world]

For complex changes go with "self-explanatory" just to fuck with peoples' confidence

[jjjalljs@ttrpg.network]

I've worked with a few people who are just incomprehensible. One refuses to write commit messages of any detail. Just "work in progress". Cast him into the pit.

There was another guy that refused to name his tests. His code was like

describe(''. () => {
  it('', () => {
     expect(someFunc()).toEqual(0);
  }
 it('', () => {
    expect(someFunc(1)).toEqual(0);
  }
 it('', () => {
   expect(someFunc("").toEqual(1);
 }
}

He was like, "Test names are like comments and they turn into lies! So I'm not going to do it."

I was like, a. what the fuck. b. do you also not name your files? projects? children?

He was working at a very big company last I heard.

edit: If you're unfamiliar, the convention is to put a human readable description where those empty strings are. This is used in the test output. If one fails, it'll typically tell include the name in the output.

[hakunawazo@lemmy.world]

git commit -m "Boss makes a dollar, I make a dime so I comment meaningful on company time"

[sukhmel@programming.dev]

To this I completely agree, a lot of people don't want to use the tools for the benefit of the future colleagues or even self

[merc@sh.itjust.works]

Also, what were you hoping to accomplish? At a minimum, are you fixing a bug? Adding a feature? Cleaning up ugly code? Trying to improve performance? Adding comments to something that wasn't obvious?

Did you change an interface that other people use in a way that might break something? Even if it's fixing a bug, is that a bug that other people might have been relying on?

I think the most problematic changes are the little fixes, because often the CL goes from something that looks like it should work, to something else that also looks like it should work. It's very helpful when the commit message describes how it was broken. Otherwise, if you have to roll back the changes you don't know what might get broken again.

[merc@sh.itjust.works]

See also semantic commit messages where you tag every commit with the type of commit: feature, fix, docs, refactor, test, etc.

My only beef with it is that they chose "feat" as a way to shorten the word "feature" when "feat" is already a word that means something different. Not every feature is a feat, and a lot of the biggest feats are actually bug fixes.

[merc@sh.itjust.works]

I get the hesitation that things can turn into lies, but that's a sign that you're doing things wrong. That also tends to happen to comments that are far away from the relevant code, like the documentation of a 100 line function. The function can change while the comment is no longer visible on the screen, so it's easy to forget to also fix the comment.

But test strings like that are designed to avoid that problem. They're right there next to your tests for a reason. You should always be right next to them when you're changing the test.

Fundamentally, this is something that has to be addressed with code reviews. If someone can commit their changes to a group repository without anybody else seeing them, you're going to get stuff like this. As soon as you get decent code reviews, you can just reject a change where there are tests without documentation, the same way you can reject a change to a test where the documentation is now out of date.

[jjjalljs@ttrpg.network]

Code reviews are important. Unfortunately, no-test-text guy convinced his whole team that he was right, and I wasn't able to block it. I'd scheduled a meeting to try to get the wider org to adopt a more sensible standard, but then there was a mass layoff

The other guy with the bad messages is at a tiny startup where they've laid off almost everyone, and the other 2 guys don't want to make waves. The CEO is big on "just ship it" (and also "why are there bugs in production? this is unacceptable!!")

[pastelkeystone@lemmy.world]

I’ve seen semantic commits done with emojis which is cute but also annoying, because they’re not as easy to type or grep for.

Semantic commits can be nice, but they can also invite bikeshedding about what’s a “feature” and what’s a “bug fix”, etc.

Not saying they aren’t nice, and if folks are using them and liking them, keep going. But if you haven’t used them before on a team, then just be aware that’s a thing than can happen.

[merc@sh.itjust.works]

I'd rather have bikeshedding over terminology that eventually results in a single word than just have free-form commits where you can never tell what the primary motivation between a commit is.