3 ms·

> From a practical standpoint, those people seem to cause vastly more trouble than whatever marginal improvements their codes of conduct might nominally capture

> From a practical standpoint, those people seem to cause vastly more trouble than whatever marginal improvements their codes of conduct might nominally capture. Do you have the ratio of the number of CoC implementers who dont add to open source vs the number of CoC implementers who do add to open source? I'm aware that there have been a couple high profile incidents where project outsiders have created drama by insisting on a CoC. But your statement could be very untrue if there is a larger number of CoCs created without incident by members of the project (I think this is likely).

wyager 7y ago
I think the actual metric we care about is (counterfactual number of problems without CoC - factual number of problems with CoC) / (factual number of problems caused by CoC) The denominator is very large (I have personally seen this come up dozens of times, including this thread), and I suspect the numerator is very small (but obviously I cannot prove it, since there's a counterfactual in there).
- nepeckman 7y ago
  I think the first term is by far the largest term in your equation. Source: the entirety of human history demonstrates that large groups of people need laws in order to function justly. In small groups, people are fine. But as you scale up, the likelyhood of bad actors reaches 1. Human societies need codified methods of dealing with bad actors, which is what a CoC does. Additionally I'd argue that you think the denominator is largest because dramatic incidents are the most visible. The vast majority of big github projects have a CoC, if they really caused so many issues you'd have way more than "dozens" of examples.
  - zozbot234 7y ago
    Not having a CoC doesn't mean you don't address bad actors, resolve disputes, etc. It just means that these things are entrusted to the project leader's best judgment. And anyone can fork the project if they think the project leader isn't doing a good job of keeping things orderly.
    - nepeckman 7y ago
      It's a matter of scale. Humans have thousands of years of history codifying rules for communities. From Hammurabi to your HOA, people have found its easiest to enforce rules when you write them down.

wyager 7y ago

I think the actual metric we care about is (counterfactual number of problems without CoC - factual number of problems with CoC) / (factual number of problems caused by CoC) The denominator is very large (I have personally seen this come up dozens of times, including this thread), and I suspect the numerator is very small (but obviously I cannot prove it, since there's a counterfactual in there).

nepeckman 7y ago
I think the first term is by far the largest term in your equation. Source: the entirety of human history demonstrates that large groups of people need laws in order to function justly. In small groups, people are fine. But as you scale up, the likelyhood of bad actors reaches 1. Human societies need codified methods of dealing with bad actors, which is what a CoC does. Additionally I'd argue that you think the denominator is largest because dramatic incidents are the most visible. The vast majority of big github projects have a CoC, if they really caused so many issues you'd have way more than "dozens" of examples.
- zozbot234 7y ago
  Not having a CoC doesn't mean you don't address bad actors, resolve disputes, etc. It just means that these things are entrusted to the project leader's best judgment. And anyone can fork the project if they think the project leader isn't doing a good job of keeping things orderly.
  - nepeckman 7y ago
    It's a matter of scale. Humans have thousands of years of history codifying rules for communities. From Hammurabi to your HOA, people have found its easiest to enforce rules when you write them down.