Many years ago I witnessed an interesting interaction on a gaming forum.
Player A was trying to recruit players for their group. Player B posted about some bad experiences they had with Player A being hard to work with in the past, which prompted Player A to reply with some vicious personal attacks. Player B then quoted Player A’s post in full with commentary to the effect of “Thanks for illustrating my point.”
When the moderator came in, they deleted Player A’s post attacking Player B, but left Player B’s post quoting it alone.
As a moderator you often know know that somebody should be removed, but for the sake of PR it's often unwise to just have an internal talk, reach an agreement that "yup, this person is an ass", and then ban them seemingly out of the blue, even if there's a bunch of excellent reasons. It's easy for drama to erupt, especially when that person has been around for a long time and is a regular.
An easy solution is to watch out like a hawk for the right incident and do it then, and sometimes to even try to intentionally push things along so that it's especially obvious to all bystanders.
And leaving some evidence to show everyone why you did it also helps.
What you describe gets far closer to why moderation is tricky. The interesting question--to me at least--is whether some subset of people can apply such reasoning consistently.
Danluu is arguing that because a set of randomly chosen people cannot agree how to apply a simple statement like "No vehicles in the park" that moderation is impossible. If so, the same would no doubt be true of the following language in an NDA:
> Each Receiving Party shall: (a) maintain all Confidential Information in confidence; and (b) exercise at least the same degree of care to safeguard the Confidential Information that it uses to safeguard its own Confidential Information (but no less than reasonable care).
Yet, US courts would not have a lot of problems interpreting this language consistently because terms like "reasonable care" have definitions that anyone trained in the law would understand. [0] The fact that uninitiated people may not be able to do so is not relevant.
>Danluu is arguing that because a set of randomly chosen people cannot agree how to apply a simple statement like "No vehicles in the park" that moderation is impossible.
I don't think this is correct, I think they're arguing that uncontroversial moderation is impossible. That moderation will inevitably lead to drama.
I think a big part of the point here is that it picked a seemingly boring, non-controversial case where you would expect agreement to be near universal but it isn't.
The game would be more interesting if the rule were "Threatening speech is not allowed."