> Do you have the same reaction to automated linters or being told to write unit tests?

There's a huge difference between technical processes that have gone through extensive peer-review processes and survived decades-long debates until a technical community reached the consensus that the practices indeed produce better software and brand-new social practices that are difficult to universally enforce. Making the statement "I don't need a list of unprofessional conduct to maintain a professional community" must be true, since you could never enumerate all unprofessional conduct anyway. You'll either leave something out (so the trolls can say "na na na, you can't get me!", in which case the method didn't work) or you'll encompass too many behaviors, which means different trolls can selectively enforce the terms of the CoC. Either way, it seems like a losing scenario.

>What about the enforcement sections of the GPL?

The terms of the GPL are criticized by many; that's why some projects choose different licenses.

> There's a huge difference between technical processes that have gone through extensive peer-review processes and survived decades-long debates until a technical community reached the consensus that the practices indeed produce better software

I'm curious - I've seen debates about test-driven development, or how to appropriately test things, but never debates about whether testing or automated linting itself produced better software.

> The terms of the GPL are criticized by many; that's why some projects choose different licenses.

Sure, but I've never heard anyone criticize it on the grounds that it insulted them or called them untrustworthy, and be taken seriously. Same with testing, come to think of it. There are lots of serious debates; "Are you calling me a bad coder" is not one of them. I'm happy to see debate about different forms of codes of conduct, but why are we taking "I am insulted by this" as a legitimate argument?

Fair point. So the response "it's insulting because it implies I'm a bad person" to a CoC discussion is shaky because the presence of a formal process designed to achieve some objective (e.g. better software, or (in the case of a CoC) a better community) does not imply any particular participant is at-fault, but rather is an evidence-based assessment that without the process the measure of some specific quality degrades. Thanks for the clarification here :)

This is totally outside the scope of your specific point, but I think it is important to note that the objective "a better community" is difficult to measure. My question: is a CoC an evidence-based response? Many software communities have moved to adopt CoCs, but what are the effects? Is there selection bias in studying a community with a CoC because (by virtue of accepting the CoC) the community is more likely to reject unprofessional behavior? How do CoCs function in a global society where actions that are appropriate in one culture might be taboo in another? Who/What is the arbiter of cultural disagreements - NAP? But the understanding of violence/aggression once again falls apart in the wake of cultural differences.

Ultimately, the answers to these questions are decided by the CoC, but who are the authors and evangelizers? The contributor covenant first appeared in 2014, and rapidly (e.g. within 1 year) saw Google (for example, there were many other large tech companies) adopt it for public projects. I believe that we should consider what the motivations are for these multi-national companies to so quickly adopt these rules.

I am not aware of non-anecdotal results here and I would be interested in seeing them - of course many CoC proponents will say that it's the right thing to do whether or not it produces a more technically meritorious project and this is legitimizing the view of a meritocracy, but I am personally expecting the answer to be that the project will in fact be more meritorious, and I think data would go a long way to arguing against (what I perceive as) biases and gut feelings against Contributor Covenant-like CoCs masquerading as data-driven concerns.