I edit my posts probably 3-5 times after making them to remove spelling mistakes, correct grammar, and make whatever changes are needed (quickly after posting).
This seems to trigger the slowdown after the second or third edit.
I also got hellband a year or two ago for submitting either a few links that made it to front-page - but then were flagged by people for political reasons, or posting a negative-Apple submit (the last non DOA submit) that might have upset a mod (titled "Brokerages Race To Feed Apple To The Muppets").
Side note: jacobian's djangobook.com is languishing and needs help. The latest commit message was a grim "Add more explicit warnings about the status of the book". See the current index.rst: https://github.com/jacobian/djangobook.com/blob/master/index...
When you submit, why aren't you told that your submissions are DOA?
Instead, you get no warning, so you believe your submissions have been accepted, when in fact nobody else can see them. If you click the "New" link you actually see your own submissions, but since nobody else can, they never get any votes.
This is quite deceptive and probably done to keep you interested in the site, until you figure out you have less "rights" than other visitors.
We're actually having an interesting conversation here about abusive flagging, in a thread I posted earlier today that was, ironically enough, abusively flagged:
For how long have you run a popular discussion web site, and what measures do you take to reduce the increased snarkiness in extended exchanges? I'd love to hear how you solve the problem.
I can't speak for that user, but I ran a forum from around 1997 to 2008 (directly involved until 2005).
Interventions helped.
This only works with moderators who aren't narcissistic, megalomaniacal and/or sociopathic (marginally or beyond). When snark got to be irritating, we would issue a written warning privately. If that went ignored, we'd revoke posting privileges for a day or two (sometimes longer if it's a bad case). Worst offenders got banned. This worked fairly well because our forum was invite only, a bit like Gmail's early days.
If we found that troublesome users were invited from the same set of accounts, we'd issue warnings to the entire tree, so to speak, or in the worst case, kill it with lightning.
Above all else, mods were forbidden from identifying themselves in any way as mods as long as they had the privilege (this will cycle on and off as needed between a select pool of users).
Sometimes they are surprisingly unaware of how hurtful their posts can be. There's a dehumanizing effect introduced when your communications are filtered through text and suddenly there isn't a "person" per-se at the other end. When this is pointed out to them, usually they understand and adjust themselves.
There are exceptions to this, of course, and the best way to deal with them is to not interact, entertain or in any way give them an audience. Hellbanning in this regard isn't my favorite technique, but it seems to work to some degree.
Now that would only be acceptable if its application is flawless. Sadly, from from what we've seen so far, it isn't.
> Interventions helped.
This only works with moderators who aren't narcissistic, megalomaniacal and/or sociopathic
Unfortunately, the Internet, startups, programming, and social media attract all seem to attract narcissistic, megalomaniacal and/or sociopathic types.
Have a set of clear rules (and not vague non-language like "don't write non-hacker, off-topic stuff") - AND ENFORCE IT, visibly. Unenforced rules aren't rules. Unclarified rules aren't rules.
Whenever someone is penalized, point them to the specific rules they violated. And keep in mind that people make mistakes every now and then. Show largesse. This isn't a production line where flawed products are tossed out at the blink of an eye.
Last, don't view everything as a "problem" to be "solved". We're talking about human beings, not Turing-complete automatons. That's a big part of the problem with how HN is run. You will never "solve" this the way you would a bug.
You need a human element to address something human. Treating everyone like a spam bot only works for spam bots, and has the side effect of pissing off the human false positives.
I disagree with the idea of having a strict code of rules. Coming up with rules and sticking to them to the letter doesn't do anything to encourage meaningful discussion.
All you end up are the trolls who do their best to toe the line while provoking other people to step over it, and the sort of people who think a conversation where everybody follows the rules yet no-one says anything of substance is preferable to a lively discussion. You end up driving away the people who just want to talk about things they enjoy or are interested in, and don't particularly care about adhering to your rigid social system.
People are people, not automatons, and the system of curating discussions should reflect that. If someone is acting in bad faith, waiting for them to violate the letter of a rule that's been written down before you do anything about it just allows that person to drive away more of your actual desired userbase.
Vague wording in rules is like legislation; they just end up as excuses for doing bad things, instead of ensuring good behaviour. "Toeing the line", if you will.
We constantly see this in trying to figure out what kind of stories are supposed to be submitted to HN.
Does it really make sense for a person like Maciej Ceglowski to be permabanned from HN? https://twitter.com/pinboard/status/111332316458135553. I think it's a shame, partly because I think he's a smart guy with a fresh view on things, but also because I don't think the reasons for banning him are very compelling.
You can't have poorly-defined rules and opaque moderation at the same time. You'll see why, when you incur the wrath of one of pg's moderation scripts for no apparent reason.
---
EDIT: As an aside, to give an idea of what a fucking pain being (inexplicably) hellbanned is, picture all your Hacker News bookmarks no longer working, because you've been banned from the entire site.
Vague wording in legislation ends up as an excuse for doing bad things because when you get enough people together to call them a government, you're bound to end up with a few bad eggs. Online discussion sites are small enough that that's not necessarily the case - it's quite feasible to maintain a core group of people who don't abuse that power.
Especially since, unlike in government, that core group of people is beholden only to whoever is operating the site, and not to a popularity contest that sees them being replaced with an entirely new group every so often. This is also why you absolutely can have poorly-defined rules and opaque moderation - the operator of the discussion site shapes what they want it to look like, and if too many people disagree with that vision then another discussion site springs up tailored to what those people want. And if the original operator is too arbitrary or capricious in their moderation, their own site dies or becomes an insular hugbox or something like that.
--
Ultimately, when people say they want stricter interpretation of the rules and more accountability on an online discussion board, what they really mean is "I disagree with some of these decisions and I want them to make the decision that I would make next time". Which is fine, but having seen a lot of the proposed alternatives in action I can't really see them as being an improvement.
Hellbanning is necessary because of the low barrier to creating a new account. You are assuming that all users are well-intentioned and can be reasoned with, but this is simply not the case. Any sizable only community will develop a number of hard-core anti-social trolls exhibiting behavior which they can only undertake online (because in real life they would get their asses kicked on a regular basis). When said trolls are also hackers you need some relatively automated and powerful tools to keep them in check. It's unfortunate that there is some collateral damage, but please understand that the choice is not between new users being inconvenienced for not, the choice is between HN being a place where civilized discourse can exist and spiraling towards a 4chan-like atmosphere.
Serious trolls know about hellbanning and how to figure out whether an account they're using has been hellbanned. The primary effect of hellbanning is to drive away well intentioned new users who are unlucky enough to break some of the unwritten rules in their first few posts.
This is all very hand-wavy. First of all, hellbanning is not always the same thing everywhere, so no, there's no guarantee that a serious troll knows how hellbanning works. More importantly, I have actually had to deal with forum trolls as an under-resourced dev for a very popular and long-running forum. It's an arms race, plain and simple. Information asymetry is to the developer's advantage. Have I gotten to the point of implementing hellbanning? No... but I understand the justification for why it's done.
How do you know this? Have you run a forum or community website, or have you been on the other side? Serious question. Must know data and provenance of data.
While this is fair question for sure not being able to solve a problem (right away) and identifying a problem are two completely different things.
Especially when asked to solve a problem as if it's a question on a test. And the fact that a problem is hard to solve or someone doesn't have an immediate answer doesn't mean there isn't a problem to be solved. Nor does it matter if they have ever run a popular discussion web site either.
In the real world to "solve a problem" such as this would take thought, time and testing all of which a commenter is not going to be able to do a good job of in a "reply" to a question posed here.
Most likely no low hanging fruit of a reply available.
I was asked a few days ago "well what would you have done then if you were Snowden?". As if I'm supposed to be able to come up with or commit to a response that makes sense on the spot. I simply won't do that.
I don't understand your question - it feels like you've misunderstood something. In case that's true, let me explain a little further.
Some time ago PG observed that deep comment threads tended to consist mostly of people getting snarky with each other. He speculated that the snarkiness might be reduced by preventing people from replying to each other immediately, without any cooling off period. To that end, and since it seemed to occur primarily on deeper threads, he instigated an exponential back-off on comments. As a thread gets deeper, so the delay increases before the reply box appears.
I don't have hard figures to hand, but I feel like the snarkiness in deeper threads has indeed been reduced, and I don't feel that the quality of conversation has diminished. One could worry that good comments aren't made because the commenter has to wait and doesn't bother, but I've seen no evidence of that.
If that hasn't answered your question then perhaps you could rephrase it and try again.
Perhaps my comment was a little snarky, but perhaps you missed the point. perhaps I was a little too subtle, or perhaps my phrasing was poor and ill-judged.
Regardless of those considerations, my question is genuine. I see a lot or criticisms here of the technical measures taken to prevent or ameliorate particular issues, expressed by people who have idealistic opinions and points of view, but with zero experience of actually dealing with the problems. That is starting to annoy me more and more, and perhaps that irritation seeped into my reply.
But the point remains. The reply back-off was implemented to help reduce the snarkiness in long threads. My feeling (with no hard evidence) is that it has worked tolerably well. My question is genuine - if this annoys kmfrk, then how has kmfrk dealt with it in the past? If kmfrk has neither experienced it, nor implemented a solution, then I feel the criticism is misplaced.
In conclusion, probably next time I won't comment.
It's surreptitious crap like this that makes HN such a pain to use for a lot of us.