> I've never seen one have this effect.

How do you think you would see it?

Under normal circumstances, what it would look like from the outside is just...someone not causing friction, because there is a code of conduct. Or someone causing friction, because there isn't one.

The former looks identical to someone who would never have caused trouble in the first place, and the latter looks, if not identical, at least very similar to someone who would have caused trouble, no matter what.

The way I see it, the people who would cause friction anyway won't be stopped by a CoC.

And the people who wouldn't cause friction are already behaving so they don't need one.

Another commenter (cannot find their name now but I saved the text) said exactly what my experience has been:

> A written code of conduct can legitimize petty disputes based on interpretations of the text, even if the issue itself is something easily addressable like using the word “master” in a programming context. It can become a tool for manipulation (pushing someone out for “CoC violation” vs simply correcting the wording).

> The way I see it, the people who would cause friction anyway won't be stopped by a CoC.

> And the people who wouldn't cause friction are already behaving so they don't need one.

Both correct. By definition, in fact.

CoCs are for the people in the middle of those. Who do, in fact, exist.

I think if someone needs to be told to be nice... it's time to grow up. Open source projects are not your mother and we shouldn't have to treat people like children, nor get chastised for having that (IMO very reasonable) stance.

It's not about being "told to be nice".

It's about setting a shared understanding of norms.

It's about whether the kinds of jokes allowed among members are "as raunchy as you please", "polite jokes only," or "none".

It's about whether you're expected to work out differences in private chats, address them in public, or bring them to a moderator.

It's about whether your comments on a code review started by another member have to be addressed by the committer before the review can go forward.

Any time you think the way you do things in the open source project (or any other kind of community) you contribute to or run is The Natural, Obvious Way Everyone Does Things, you're probably wrong, and some other people or projects do things exactly the opposite way and think that's just as natural and obvious.

Presumably the code of conduct was introduced to solve a problem. So I'd be interested in hearing about an example of a community that had problems with their social fabric, and then introduced a code of conduct, and then could demonstrate that those issues in their social fabric were addressed.

In the history of the world? Yeah. Anytime you have a communal living space, you're going have to set norms on what constitutes shared community stuff. Is the milk in the fridge George's and no one is allowed to touch it, or do we, as a household, want milk communism and it just goes as shared groceries and we go often enough that there's always milk? There's not a right answer, but it's up to the community to decide what's normal in its culture. People come from different backgrounds, were born in different countries and have totally different expectations as to what constitutes "don't be an asshole". Do you put your elbows on the table? Is slurping your soup a sign of respect or disrespect? What you don't see is people who see a lack of a code of conduct and then just don't join.

Don't these tend to be discussions between the people involved?

Whether or not anything needs to be written down afterwards seems like it would be a separate thing.