this post was submitted on 27 Jun 2024
425 points (95.5% liked)

Programmer Humor

19444 readers
33 users here now

Welcome to Programmer Humor!

This is a place where you can post jokes, memes, humor, etc. related to programming!

For sharing awful code theres also Programming Horror.

Rules

founded 1 year ago
MODERATORS
 
you are viewing a single comment's thread
view the rest of the comments
[–] [email protected] 45 points 4 months ago (2 children)

Please use Conventional Commits. Simple and easy to use. Plus it is very easy so combine with Versioning techniques like Semantic Versioning.

[–] [email protected] 11 points 4 months ago* (last edited 4 months ago) (1 children)

Honestly, I've worked with a few teams that use conventional commits, some even enforcing it through CI, and I don't think I've ever thought "damn, I'm glad we're doing this". Granted, all the teams I've been on were working on user facing products with rolling release where main always = prod, and there was zero need for auto-generating changelogs, or analyzing the git history in any way. In my experience, trying to roughly follow 1 feature / change per PR and then just squash-merging PRs to main is really just ... totally fine, if that's what you're doing.

I guess what I'm trying to say is that while conv commits are neat and all, the overhead really isn't really always worth it. If you're developing an SDK or OSS package and you need changelogs, sure. Other than that, really, what's the point?

[–] [email protected] 0 points 4 months ago

You can always water it down. The point is to have some order in the commits. Otherwise is just messy.

[–] JackbyDev 4 points 4 months ago (1 children)

Any standard that wastes valuable space in the first line of the commit is a hard sell. I don't see the point in including fix/feat/feat! just for the sake of "easy" semantic versioning because generally you know if the next release is going to be major or minor and patches are generally only only after specific bugs. Scanning the commits like this also puts way too much trust in people writing good commit messages which nobody ever seems to do.

Also, I fucking hate standards that use generic names like this. It's like they're declaring themselves the correct choice. Like "git flow".

[–] [email protected] 1 points 4 months ago (1 children)

You can always adapt to your how repo. But yeah, that's the point. If you can trust people to make changes on a repo then you should be able to trust them in using some kind of commit structure.

Generic names are probably used in order to crate a familiar, easy to remember, structurized commit format.

[–] JackbyDev 0 points 4 months ago

The generic name I'm complaining about is "conventional commits", not "fix" and "feat"