Comment by jchw

So, I definitely believe this story 100%: when I was on the anglosphere Internet in 200x, there was a lot of elitism and hazing rituals of sorts, among other things. It was a very real and unfortunate thing that coincided the otherwise excellent experiences (IMO) of being online at that time.

Still, I really don't think most people need to be told not to tell other people to kill themselves, and in many places where I hung out when I was younger I strongly believe you would have been tempbanned for "flaming". I was a forum moderator and I can tell you I would not have hesitated.

But you said the magic words, so it bears addressing; I think we all get the picture that the Code of Conduct drama usually doesn't have much to do with the actual rules that are contained within, which really aren't that controversial on their face, but rather the way in which power is moved from stakeholders within a project to other people by virtue of initiatives like establishing Code of Conducts and the governance structures that enforce them. And, I think most people will probably not get upset over the idea that telling someone else to go kill themselves might get you suspended from a discussion forum... Rather, the drama comes in when you see the reach of a project or organization's CoC start to extend outward past what people actually want to stop (toxic, unproductive communication) and past the edges of the project (and into policing the rest of the Internet.) Two notable examples I'd cite are Python with Tim Peters (who as far as anyone can tell genuinely didn't do anything wrong) and Freedesktop.org with Vaxry (who can be a bit immature, but is primarily accused of not moderating the Hyprland Discord... Which is a fair complaint about the Hyprland Discord, but not a very good reason for him to be banned from Freedesktop.org.)

Of course, truthfully, there is no 100% winning answer here; if the stakeholders who have control over a project by virtue of being the original developers don't want to cede any control to people for CoC enforcement, they don't really have to (although in reality, external pressures to implement one might make it an untenable position to hold.) In that case, you have to rely on those people to hold themselves accountable to reasonable conduct, and nobody's perfect. It's kind of like when police departments conduct internal investigations and find no problems; even if you're pretty pro-police, you must feel somewhat skeptical that they actually were reasonably impartial in conducting said investigation.

But, I generally side with The Evil I Know, which is that the project authors and biggest stakeholders should generally maintain most of the power and control in an open source project including the ultimate decisions regarding moderation. In cases where developers have proven particularly egregious with their conduct, forking has proven to be effective enough as a mitigation strategy, and the fact that it comes at a cost is a sort of feature, as it's better if a power shift like that isn't easy; while I can't guarantee that the original authors and maintainers of a project will act reasonably and impartially, I can at least say that I expect them to have the project's best interests at heart, whereas the kinds of people that go around looking for established projects and organizations to join roles that have authority tend to not be the kinds of people you usually want in those roles. Having it be difficult means you need people who genuinely care about the project rather than the types of people who just kind of seek power. (And I am sorry, but there are fuckloads of those people among us and they are absolutely dirty enough to hide under the guise of anything to get a modicum of control. Running an online community for any appreciable amount of time opens your eyes to this IMO.)

All of this to say, it reflects poorly on the state of the Internet at the time and KDE's mailing lists that the situation happened and was possibly not rectified in a way that is satisfactory (it sure doesn't sound like it.) The correct thing to do is obviously to issue a ban, and you don't need a rule book of any kind to figure that out. I think when people push for these things during major incidents, it's misguided at best, because usually the core problem was not that a "don't tell people to kill each other" rule didn't exist, but that people actually would've needed such a rule to decide the behavior was unacceptable in the first place. This isn't some complex gray area case. I don't think people are acting in bad faith when they suggest it as an option after a drama incident, but I still think it's the wrong knee jerk 99% of the time.

(The most favorable thing I can say is that I think a CoC might possibly have value in very large projects like Linux or Kubernetes, but so far the execution has always felt like it leaves something to be desired. Seeing people occasionally openly threaten to contact the CoC committees over effectively technical disagreements leaves a bad taste in my mouth.)