Should we use Conventional Commits?

[tcharding]

We have had various discussions/attempts at introducing tooling to help create changelogs, as a step towards that should we start using Conventional Commits?

[apoelstra]

Yeah, I'd be willing to give it a shot. Since none of us are familiar with it I guess we'll have to try policing each other.

[tcharding]

chore

[sanket1729]

Not familiar with it. But open to experiment as a soft(non-enforcing) policy.

[dpc]

Warning a while, you'll be writting chore: in front of everything.

[tcharding]

I hit that already, and I'm not super keen on calling arbitrary improvements 'chores', I love programming but I'm not that fond of doing chores - seems incongruous.

[tcharding]

I guess you guys are using it in Fedimint @dpc - in hindsight was it a good idea?

[dpc]

We don't have the will to be more careful about chore/feat/refactor distinction, even less so to scope field, so I feel like most stuff is just a chore, and all in all it doesn't really give us anything, other than just forcing lower case letters.

We might need to up our game in the future, but RN there's so much fundamental and important things to get done quickly, that stuff like that is not very important.

On the positive side - it's not a lot of effort, and doesn't feel cumbersome.

[tcharding]

I've already found the fix: and feat: prefixes useful in enabling looking back at a list of patches to see what semver change needs making when doing release PRs.

Until we work out exactly what tool we are going to use to autogenerate changelog notes I can't see the benefit in putting chore: before everything.

[tcharding]

Note that that is also what our github labels are meant to do but I find it easier to look and the git index in the terminal that using github.