互联网与文化danluu.com
Hacker News 评论被低估了
HN comments are underrated
2016 — 2021，它一次次回到首页
黑点是上首页（≥ 50 分）的那次投稿，空心点是没冲上去的投稿，点越大分数越高。
导读
Dan Luu 在这篇短文中为 Hacker News 的评论区正名：尽管 HN 常被批评刻薄、傲慢，但它的评论质量实际上被低估了，很多领域专家会在评论中分享一手经验。文章鼓励人们认真写评论，甚至把好评论扩展成博客文章。
为什么经久不衰：它讨论的是社区评论质量这一永恒话题，多年来反复引发 HN 用户的自我反思，对任何参与在线讨论的人都仍有启发。
HN 评论精选
好内容写成博客比埋在评论里能获得多得多的曝光和反馈
我附和 Dan 多写博客的建议。我只是一个很普通的 HN 评论者，虽然写评论时尽力保持礼貌、分享经验，但我不是 Alan Kay，我的评论肯定上不了“HN 精选”。上周 Salesforce 放弃收购 Twitter 的消息传出后，我本想写条评论谈谈 Twitter 该怎么办，结果写得太长，索性扩成了 Medium 文章。让我惊讶的是，这篇文章获得了 2.8 万次浏览和 755 个推荐；如果只是写成 HN 评论，大概只能得到 5-10 个赞，还可能引来一串说我不切实际的讨论。也许可以有一个介于 HN 和 Medium 之间的网站，让你以极低的摩擦把评论扩展成博客文章？
pavlov查看英文原文
I second Dan's advice of blogging more. I'm a very average HN commenter. I do put in effort in writing here, trying to be civil above all and sharing my experience where it could be of interest. But I'm not Alan Kay, I've never rewritten a distributed deep learning system in Haskell using a genetically optimized Paxos consensus protocol, and my entrepreneurial experience is a loose string of "don't do this" case studies at best... So my comments certainly won't make anyone's "Best of HN" list. Last week, after the news broke that Salesforce walked away from buying Twitter, I was about to write a HN comment about what Twitter could do. The text got long enough that I decided to expand it into a Medium post instead: https://medium.com/swlh/twitter-could-be-the-next-mozilla-e788e3bfd841#.x0gbzodo4 To my surprise, the post has 28,000 views and 755 recommends so far. If I had written it as a HN comment, it would have got maybe 5-10 upvotes and perhaps spawned a short discussion thread about how unrealistic my idea was. (Please don't bother to criticize the content of the blog post in replies here -- I'm just using it as an example of blog vs. comment.) I love reading HN discussions... But maybe there could be a site that slots between the HN and Medium formats, and lets you expand your comment into a blog post with minimal friction? Call it "HN Long-Form" or whatever. Ideally it would interface with the HN comment system so that you could mark your comment with something like "Promote to long-form" after you've written it. That would create an editable post on the long-form site. You could then later expand your comment there, and publish it on the long-form comment aggregator site. (Maybe I should just build this myself and see if it feels right.)
HN 的批评并不比其他行业的评审更苛刻，关键在于学会过滤反馈
再读一遍，我不太同意 Dan 文章开头的说法，他似乎在说 HN 很糟糕，只有少数明星用户的评论出淤泥而不染，我觉得这不公平。是的，HN 确实有刻薄挑剔的刻板印象，但这忽略了两个事实：其一，HN 上的批评并不比其他创意行业从业者承受的更糟——我上过艺术设计院校，学生甚至老师的点评绝大多数时候比 HN 风格更严厉，也未必更有用；想想初出茅庐的小说家被报纸评论员批得体无完肤，你能做的也只是咽下去继续写下一本。利用反馈的关键在于过滤和提炼，没有人永远正确。
pavlov查看英文原文
On second read, I'm not sure I agree with the first paragraphs of Dan's post at all. He seems to be saying that HN is terrible, but a handful of comments from star posters rise above the muck. I just don't think that's fair. Yes, the cliché is that HN is a place full of mean, entitled semi-autists who will criticize your site's CSS whitespace formatting when you ask for business feedback... And of course there's a grain of truth to that (persistent stereotypes usually don't come out of thin air), but it misses the mark on two dimensions. The first is that the criticism you get on HN is no worse than what other aspiring creative professionals suffer. I went to an art and design college, and the critique you'd get from students and even teachers was 99% of the time harsher than the HN style, yet no more guaranteed to be useful. Consider a first-time novelist who spent years on a book. One day it gets critiqued in a newspaper. The professional critic might find that the author has a clumsy style, poor research, paper-thin characters, and seems to lack the life experience to even write about the topic. What do you do after that kind of criticism? You suck it up and go back to work on the next novel. Making use of feedback is all about filtering and reducing multiple sources into something actionable. Nobody is right all the time. Your parents were wrong. Your teachers were wrong. Your peers were wrong. Your professors were wrong. Your boss was wrong. Your cofounders were wrong. Your investors were wrong. HN commenters were wrong. Still it's worth taking in all these inputs as much as you can. The other dimension of HN comments is that they can be surprisingly deep. When an arts or culture topic makes it to the front page, it seems like someone comes out of the woods with the perfect personal anecdote. Whether it's Mondrian, Messiaen or Modiano, there's always someone on HN who happens to have a passion for it. HN comments are underrated, but it's not just because of star power: it's everyone's contributions that make it consistently worthwhile for me.
数据表明 HN 的集体判断力在下降，可通过投票加权和折叠长分支来改善
我分析 HN 帖子数据时发现，几年前用户曾准确预测过哪些 Show HN 项目后来拿到了融资，这些项目获得更多赞；但近年的数据中这种关联消失了，似乎用户涌入降低了判断质量。一种改进办法是按用户在 HN 的资历给投票加权。另外，早期评论仅凭时间因素就压制了后来的评论——首页帖子一小时内就有 30 多条评论，后来者只能对着虚空发言。可以把长评论分支默认折叠，每个分支只显示 2-4 条。
anton_tarasenko查看英文原文
When I worked with the HN post data, I noticed that some years ago HN users had correctly predicted the "Show HN" projects that later got funding. Those projects had more upvotes. The more recent data has no such connection. It seems that the influx of users reduced the quality of judgement. So one way to improve HN submissions and comments is to weight points by the user's tenure on HN. I also suspect that early comments dominate late comments by the time factor alone. The sorting algo gives a brief advantage to new comments, but old comments are more visible. A post on the front page gets 30+ comments in the first hour, and latecomers can only post into the void. To address that, long branches could be collapsed by default, leaving only 2-4 visible messages per branch.
引用 PG 的分析：刻薄可以用规则约束，愚蠢则更隐蔽也更难治理
PG 的文章《我从 Hacker News 学到了什么》说得很到位：坏评论比坏帖子更难治理。评论的坏主要有两种——刻薄和愚蠢，二者高度重叠但应对策略不同。刻薄容易控制，立规矩并执行即可；愚蠢更难对付，因为愚蠢的人往往意识不到自己愚蠢。最危险的坏评论不是长篇的错误论证，而是愚蠢的段子。评论质量与长度有很强的相关性，平均长度是衡量社区评论质量的好指标——愚蠢更多表现为想法太少，而非想法错误。
wiggler00m查看英文原文
PG's essay "What I've Learned From Hacker News"[0] is on point: COMMENTS: Bad comments seem to be a harder problem than bad submissions. While the quality of links on the frontpage of HN hasn't changed much, the quality of the median comment may have decreased somewhat. There are two main kinds of badness in comments: meanness and stupidity. There is a lot of overlap between the two—mean comments are disproportionately likely also to be dumb—but the strategies for dealing with them are different. Meanness is easier to control. You can have rules saying one shouldn't be mean, and if you enforce them it seems possible to keep a lid on meanness. Keeping a lid on stupidity is harder, perhaps because stupidity is not so easily distinguishable. Mean people are more likely to know they're being mean than stupid people are to know they're being stupid. The most dangerous form of stupid comment is not the long but mistaken argument, but the dumb joke. Long but mistaken arguments are actually quite rare. There is a strong correlation between comment quality and length; if you wanted to compare the quality of comments on community sites, average length would be a good predictor. Probably the cause is human nature rather than anything specific to comment threads. Probably it's simply that stupidity more often takes the form of having few ideas than wrong ones. Whatever the cause, stupid comments tend to be short. And since it's hard to write a short comment that's distinguished for the amount of information it conveys, people try to distinguish them instead by being funny. The most tempting format for stupid comments is the supposedly witty put-down, probably because put-downs are the easiest form of humor. So one advantage of forbidding meanness is that it also cuts down on these. Bad comments are like kudzu: they take over rapidly. Comments have much more effect on new comments than submissions have on new submissions. If someone submits a lame article, the other submissions don't all become lame. But if someone posts a stupid comment on a thread, that sets the tone for the region around it. People reply to dumb jokes with dumb jokes. Maybe the solution is to add a delay before people can respond to a comment, and make the length of the delay inversely proportional to some prediction of its quality. Then dumb threads would grow slower. [0] http://www.paulgraham.com/hackernews.html
历次讨论
同类还值得读
如今电脑上几乎所有操作在感知上都比 1983 年更慢（2017）
Almost everything on computers is perceptually slower than it was in 1983 (2017)
2017 年的推文串认为，尽管计算机硬件比 1983 年快得多，日常操作的感知速度却常常更慢：启动、界面响应和网络应用被层层抽象、动画、后台检查与网络依赖拖累。
互联网与文化上首页 5 次 · 2017–2020twitter.com网站肥胖危机
The Website Obesity Crisis
Maciej Cegłowski（idlewords）的经典演讲文稿《网站肥胖症危机》，批评现代网页越来越臃肿：动辄数 MB 的 JavaScript、追踪脚本和广告，把本可以轻…
互联网与文化上首页 5 次 · 2016–2021idlewords.com编辑精选亚马逊那本标价 23,698,655.93 美元的苍蝇书
Amazon’s $23,698,655.93 book about flies
博文讲的是亚马逊上一本关于果蝇的二手书被标到 23,698,655.93 美元。
互联网与文化上首页 5 次 · 2011–2022michaeleisen.org浏览器 User-Agent 字符串的历史
History of the browser user-agent string (2008)
2008 年的经典文章用戏谑的“创世纪”口吻，讲述浏览器 User-Agent 字符串如何从简单的标识演变成层层叠加的伪装：浏览器为了绕过按 UA 分发的网站检测，纷纷把自己伪装…
互联网与文化上首页 5 次 · 2012–2022webaim.org编辑精选最后的问题——艾萨克·阿西莫夫
The Last Question - Isaac Asimov
阿西莫夫最著名的短篇科幻小说《最后的问题》的在线全文。
互联网与文化上首页 5 次 · 2009–2019multivax.com编辑精选