5 ms·
HN has been under downward quality pressure for a long time. Figuring out how to withstand it has been the core idea all along: HN is an experiment. As a rule,
by dang 3y ago
HN has been under downward quality pressure for a long time. Figuring out how to withstand it has been the core idea all along:
HN is an experiment. As a rule, a community site that becomes popular will decline in quality. Our hypothesis is that this is not inevitable—that by making a conscious effort to resist decline, we can keep it from happening. - https://news.ycombinator.com/newswelcome.html https://news.ycombinator.com/newswelcome.html
pg wrote that over 15 years ago. I've been saying for (a mere) 10 years that we're trying to stave off the arrow of internet entropy: https://hn.algolia.com/?dateRange=all&page=0&prefix=true&query=stave%20off%20by%3Adang&sort=byDate&type=comment https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que... – not doable forever, no doubt, but doable for a while if we expend enough energy.
Based on experience so far, the way to do this is through a combination of a dedicated community, software that does what software can, and moderation to bump the system out of its failure modes.
Trying to hide from the outside world is mostly not the way. pg used to do tricks like Erlang Day on occasion, but having that as the main strategy would be like trying to avoid infection by never going outside. Far better is to have a robust immune system, if possible. Trying to avoid (resistible) pathogens weakens the immune system, and hiding HN from new users is a path to decrepitude. The latter is probably a greater threat IMO, and too easy for established users to discount.
Spam, and its cousins like content marketing, could kill HN if it became orders of magnitude greater—but from my perspective, it isn't the hardest problem on HN. That's because of the dedicated community, which flags these things, reports them when they escape in the wild, and is vigilant about quality. Without such a community, HN would have died long ago.
By far the harder problem, from my perspective, is low-quality comments, and I don't mean by bad actors—the community is pretty good about flagging and reporting those; I mean lame and/or mean comments by otherwise good users who don't intend to and don't realize they're doing that. There's an unholy dynamic between those and the upvote system, so worse comments often get upvoted more than better comments do—often enough to choke the threads with weeds. That's the high-order bit and what I spend more time worrying about—not Cassandra E Oakley and her trading system*, nor the latest startup voting ring and whatnot. If those ever become the high bit, we might be doomed, but we should see it happening long enough in advance to readjust.
p.s. (@dang doesn't work - I only saw this by accident. Well, not by accident because it ended up at the top of the thread)
* https://news.ycombinator.com/item?id=39425371 https://news.ycombinator.com/item?id=39425371 - unkilled for the occasion
- berniedurfee 3y agoHave quotas ever been tried or considered? Like users being allocated a few upvotes per day or week? I wonder if that would give enough pause for users to consider if a comment is worth their upvote.
- dang 3y agoWe haven't tried it. We've considered it, but not deeply.
- throw10920 3y agoYou make very good points, as expected. However, while I completely agree that spam isn't anywhere close to being a threat to Hacker News, I'm not so sure about subtle, well-executed content marketing/influence campaigns. By design, these are much less overt than mere spam links to a porn website, and there are plenty of places for them to pop up on HN - for instance, in any number of recent threads lamenting the sad state of modern appliances/tools (like vacuum cleaners, e.g. [1] - and it would be extremely difficult to even approximate the scale of the problem. I'm not advocating hiding from the outside world in the same sense that one might try to intentionally isolate their community to prevent others from joining - just in the sense of avoiding a specific discovery mechanism (Google) that results in extremely strong incentives for content marketing companies to infiltrate and influence. More specifically, I posit that the amount of organic discovery and community growth that we would get would be significantly outweighed by the attention from content marketers - most people that I know discovered HN through blog posts and individual recommendations, and not through search engine results. Ultimately, of course, it's up to you - but please think about this issue (and particularly the tradeoff of the upsides vs downsides of being highly-ranked in Google results), and look for indicators of content marketing. Hacker News has, by far, the highest quality-size product of any community that I've seen on the internet ever, and I would dearly love for it to stay that way for many decades to come - which I don't think will happen if "Hacker News product recommendations" becomes a currency with supply comparable to GitHub stars[2]. [1] https://news.ycombinator.com/item?id=39238010 https://news.ycombinator.com/item?id=39238010 [2] https://news.ycombinator.com/item?id=36151140 https://news.ycombinator.com/item?id=36151140
- 3y ago