No doubt it's the worst solution, except for all the others. We have few software tools for dampening the descent of this place into flamewar. If HN burns itself to a crisp, what good does that do anyone?
I do think it can be improved by making it more of a probation system that explains what's happening and how long the probation period is, presumably with some sort of exponential backoff so that over time probation converges to a permanent penalty. This is on the list to implement.
We rate limit accounts when they post too many low-quality comments too quickly and/or get involved in flamewars. If an account is doing a lot of things that damage HN for its intended purpose, obviously we have to do something.
> If you want post limits, enable them for everyone.
That would be inconsistent with optimizing for curiosity [1]. There are heaps of users from whom the more comments HN is lucky to receive, the better.
It would be great if admins had a way to tag the "low-quality" comments that triggered the rate limiting, so people can kind of learn by example. Like "Please slow down and stop posting things such as <link>, <link>, and <link>."
This rate limiting seems to happen to me quite a bit, and when I've E-mailed, it's "Hmm, you look like you're posting fine now. We'll remove the limit." Leaving me scratching my head, searching in vain through my comment history for anything that might have plausibly triggered the action.
I agree with your first point. People learn from feedback, so it would be better if we could give more precise and structured feedback. This is a longer term goal.
It's hard though because everyone who asks a question wants a detailed answer and detailed answers take a lot of time and energy, adding up to way more time and energy than we have available, or could ever have. We do our best, but it's a hard constraint problem.
What's the purpose of links like that in your moderation comments? I click on them assuming they'll lead to additional explanation, but they're only ever showing that you've used some particular word or phrase before.
I'm not sure what I'm supposed to be taking away from that.
They do lead to additional explanation—usually lots of it. For example, if you follow the link I just mentioned (here it is again: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&sor...), is it really not obvious what the point is? The point is that we try to optimize HN for curiosity and that this has interesting consequences. You do have to scroll back through the search results to find the good explanations, but that's not so hard.
One thing I intend to do eventually is compile those past explanations into a set of commentaries that could be linked to in the future. For now, the genres through which I explain this stuff are (a) HN comments and (b) emails, and I can't link to emails, so you get past HN comments.
It's really no better than providing a link to a google search when someone asks for an explanation. You'd call that unsubstantive when a user did it, and it's unsubstantive when you do it.
I don't think that's fair. A google search isn't necessarily unsubstantive—it depends—and the HN Search links I post are all pointing to consistent explanations.
I do think it can be improved by making it more of a probation system that explains what's happening and how long the probation period is, presumably with some sort of exponential backoff so that over time probation converges to a permanent penalty. This is on the list to implement.