I've realised that I've built up a set of heuristics for deciding what to read on HN:<p>- don't click on anything with the word "quantum" in it — it's either too technical for me (physics or computing) or mainstream fluff with no substance<p>- don't click on very specific programming language links, unless they're about Python or frontend web stuff (just not interested in languages I don't use)<p>- will read anything from certain domains — danluu.com, stratechery.com, wikipedia.org, fermatslibrary.org<p>- won't click on the latest iteration of "ML tutorial for beginners" that makes it to the front page (not the right audience, but nice to see this stuff getting popular)<p>What are your heuristics?
Hacker News is basically slowly but surely converting me more into a technical person. When I came here first, I mostly looked at the psychological articles and some web articles. But when I felt a bit more focused and bold I'd go for more technical articles. I have been doing this ever since and I have noticed there are two types of technical articles I really like:<p>1. any computer graphics reverse engineer article, or reverse engineering regarding a game console, or the posts about Dolphin the gamecube emulator.<p>2. any security article. Especially about exploits or vulnerabilities.<p>I was surprised at how well I understood them. But I think in part it has to do with that I sometimes venture out and expand my comfort zone. I keep on expanding my comfort zone through Hacker News, it really supplemented my computer science education quite well.<p>I also use it a lot as a search engine to find high quality educational content. All those upvotes matter. And if the upvotes don't matter -- the educational resource is quite bad -- then there's always some insightful comment about a good resource. I learned a thing or two about deep learning this way and how it relates to topology. I've never read anything about topology before! That's just really cool that some resources can give you a basic intuition about it.
I've found that when there are a decent number more comments than upvotes (say 1.25+ comment / upvote ratio) it is generally a controversial topic, probably with a huge amount of subcomments under the top comment. If it is in my domain, I tend to ignore it, as I've often heard both sides many times before and its draining to see debates play out again. If it isn't in my domain and it is interesting, I'll check it out and often learn something. If its outside my domain and not a particularly interesting topic, if i'm bored i may check it out to entertain myself for a bit<p>for Ask HN, the comment / upvote ratio seems like it indicates a topic that people are interested in and want to share their two cents (like this one), rather than a controversial one<p>i also click on most links from wikipedia, as any wikipedia entry that gets to the front page seems like it must be very interesting / esoteric. however they turn out generally to be hit or miss
I almost never click through on the source link initially. The only times I do are if the topic is something <i>very</i> specifically related to my personal interests / work, and/or from a reputable domain, and possibly only if it has a lot of upvotes. I'd say that links to github.com, and arxiv.org are among the top ones that I'm likely to click through to immediately.<p>OTOH, anything that looks even vaguely interesting gets an upvote, since that is the easiest way to "save" a HN link for later (the "favorite" feature requires you to at least click through to the discussion page first, which is just enough friction that I mostly don't bother).<p>Anything in between those two extremes, my heuristic is to click through to the discussion page, spend some time reading the comments, and then decide if the source article is worth reading or not. I feel like you can usually tell a lot about the quality of the underlying article based on what kind of discussion it generates.
1. anything crystal / ruby related<p>2. rust and golang related stuff, so I can further justify my choice of crystal<p>3. security exploits<p>4. dramas, like angry co-founders duking it out on HN, love that stuff<p>5. philosophy of mind / consciousness related stuff<p>6. stuff that tells me when I can get a consumer-grade quantum computer (never happens, or can only multiply 3 * 5 for 50 million dollars)<p>7. anything that sounds like a horizontally and vertically scalable database solution -- been looking for something like citus or google cloud spanner, but with a $5/month starting plan since forever ago. Note: a $90/month starting plan != scalable imo, as it doesn't allow me to justify developing with it.<p>8. "here is a super fast hash table implementation"<p>9. "new data structure" (happens almost never)<p>10. "programming language performance benchmark"<p>11. microsoft/google/facebook/apple/amazon hating never gets old<p>12. news that [language I care about] now runs on web assembly (never happens)<p>13. "RSA encryption broken by efficient factorization of large semiprimes" (world would end for a few days)<p>14. proof or disproof of p = np<p>15. autonomous driving stuff<p>things to avoid:<p>1. anything related to react/vue, though I might click if it's vue just for them sticking it to react<p>2. anything about VC stuff, because I don't come here to read that<p>3. news that [language I don't like] now runs on web assembly (every day)<p>4. x new javascript framework
I click on most things that start with "10 ways to..." and then immediately regret it. I then repeat the process with "Kubernuts (is/isn't) the best thing ever" and then immediately regret it. And then I click on the comments of any article that I have industry specific knowledge on and tell everyone why their opinion is wrong without reading the article.<p>Gosh I need a new way to kill time.
- Anything with >500 upvotes that isn't general news, the more technical/esoteric, the better.<p>- "How to write/build a [something]" if that "something" is fairly low-level, like a compiler.<p>- Just about anything involving bees.<p>- Highly-voted threads with mundane headlines -- i.e. either the content is good enough to not need a catchy headline, or its headline has been changed to be less clickbaity -- which might mean the mods thought it was worth saving rather than flagging into oblivion.<p>- Virtually any Show HNs that happen to make it to the front page<p>- Ask HNs that have > 50 comments.
Click on:<p>- stuff I know a bit about, and can perhaps offer some anecdata on rather than commentary (because you're all far smarter than me on everything, and that's wonderful)
- stuff that reminds me that working for all these hours is worth it after all (because I need it)
- stuff that tells me that working for all these hours is preposterous and I should re-assess what I'm doing (because I need this, too)
- Ask HN posts that show some human humility and a wanting to learn
I wrote a simple extension for hckrnews.com to hide submissions based on words and domains: <a href="https://gist.github.com/ivmirx/66a0015884d44297ea05a8c54d93566d" rel="nofollow">https://gist.github.com/ivmirx/66a0015884d44297ea05a8c54d935...</a><p>I filter out all the languages and tech that I don't use, corporate news, US news, crypto news, and some bad journalism.
If you want pure technical crunchy hits of great goodness (no pol no fluff no bs) the best heuristic I've developed is to look for the things that reach front page with 30~40 pts but have almost no comments. These usually tend to be PDFs of research papers describing some awesome but highly advanced or technical cool thing, right on the edge of comprehensibility.<p>My theory is, we're all crouched around like proto-humans before the Monolith, upvoting and hoping someone bold and clever will touch it, or better yet, that one of the aliens who put it there will show up and tell us what it is. ;-P
I start by scanning through the front page and clicking "HIDE" on:<p><pre><code> Betterridge Headlines
Articles with a (YEAR)
Articles with a [PDF]
"Do X in Language"
Blockchain-related
Mac-related
"Ask HN: List things"
Hiring notices
</code></pre>
- Specific domains:<p><pre><code> All newspapers
All longform sites
All clickbait sites
All health information
Medium
Github
Wikipedia
Stratechery
</code></pre>
This usually clears out all the time-sinks, ads, "Github spam" and clickbait. And sometimes you find a Tesla article or 2 that's been pushed back to the 2nd or 3rd page.
I've basically learned web design and development through HN.<p>The only links I click on are related to making static websites without any frameworks. Otherwise, if the websites domain is a new one, I'll click on it to see if the website has an interesting font or layout.<p>I never check the comments on HN for many things non-STEM, more specifically economics, philosophy, psychology, education, academic research, political science, fashion, and more. Those comments are generally filled with people trying to use first principles poorly to speak on subjects they don't have industrial experience in.<p>The best resources I've gained via HN<p>1. practicaltypography.com - Basically my foundation when it comes to designing any text in personal documents, websites, video games and more.<p>2. worrydream.com - The best HCI resource freely available on the internet. It opened my eyes to the sorry state of software with respect to usability.<p>3. Data Oriented Design - Research into this field gave me a definitive answer on the merits of continuing to struggle with learning functional programming versus going all in on traditional C style development.
- Golang.<p>- “X is dead/dying” just because I want to see if there’s any sort of logic to it and to see the comments (which can actually be pretty useful)<p>- Psychology-related things (especially about depression, PTSD, burnout)<p>- Tutorials that I may find useful<p>- Security/exploits (like the latest ESLint issue)<p>- Apple-related articles (because I use Apple products but know they aren’t perfect and want to see both sides)
A post with a few comments ranks much higher for me than one with just upvotes. The main reason I upvote something is because I want it to have more exposure so more people will comment on it. Having an expert in the field or the author of the paper etc just pop in and start answering questions is my favorite thing about HN.<p>Unsurprisingly, I check the comments first, unless it specifically piques my interest (which isn't hard), and then I leave the comments open as a parent tab to check later because if I'm interested in a topic I'm interested in opinions of it.<p>These are more likely for me to click through to:<p>* Databases and programming languages: theory, standards, etc.<p>* Specific languages I like reading about: rust, go, .net<p>* A little bit of general world news<p>* Psycology<p>* Security<p>* Anything really technical. I like picking up on new modes of thought even if I don't understand the topic.<p>* Anything that stays on the homepage for a while I'm likely to check out
I tend to (as others have mentioned) focus on languages or fields I'm familiar with. Lots of articles on frontend stuff, and JS libraries, even ones I don't use. Mainly just to get a feel for what is out there. A lot of the time I like to read comments more than the actual article on a lot of things, because I find opinions on HN (specifically comments that have long threads with multiple replies) to offer a better understanding of things than the actual article. A lot of articles are somewhat biased, intentional or not, and reading the comments (most of the time) offers a better perspective from people actually using/working in/experiencing, things the articles are talking about.
When in doubt from the title:<p>unknown domains, blogspot.com, some known high-quality sites > Github > general news/media sites > medium.com, hackernoon.com, dev.to<p>I tend to look at Show HN. I look at the new queue a lot, less so at the front page.
I don't consciously or systemically apply any heuristics, but if I think back on the topics that tend to pique my interest: Business and VC, economics, infosec, tech politics, FL/OSS drama, psychology / human nature, and oddities that seem out of place here (because those stories must be especially interesting!). Occasionally web dev tools if they look like something I might use — for example, new static site generators.<p>I nearly always read the comments first and rarely click through to the original story. The comments are the point of HN for me.
I find myself thinking about whether the comments will have anything interesting in them. If I think there will be then I click on the comments page and read through the top few comments. If there's a lot of discussion about the OP link itself only then will I actually click the link. I do the same for reddit.<p>Side note: I think that there are many articles/blogs on the internet that have less information than a good reddit/HN comment and this is why many users of HN and Reddit go to the comments first.
Posts that are either Ask/Show HN, or links to articles/tutorials of tools and services of personal usage - personal backup articles, a nice little desktop/mobile app, a new messaging service of note, or a new mobile or laptop that is trying to something new etc. Comments are where I look for interesting content on such posts.<p>I almost always avoid those yet another JS and JS/Native/web/blogging/static site framework articles. Android dev, my bread and butter, is rarely shared and discussed on HN and since I don't (usually) browse HN during work I would anyway be not interested in things related to my work (I try to compartmentalise it like that - leave work when I leave office). Most of the tech discussions here are of backend, dev ops, and web dev which don't really interest me. I also avoid discussions or articles which have low level implantation details.<p>I also avoid those <i>list</i> articles - "best way(s) to….", "best/most X..", "Y is X…" etc. Or an article that sounds like a rant from the title unless it's by someone like DHH or the Pinboard founder or has a large number of comments. Or posts about something from my country or this side of the world.<p>I also look for one of those rare excellent non-tech article (preferably a long-read) - on policy, on history et cetera. And articles related to policy, privacy etc - comments are pretty informative on such posts, even though the posts are overwhelmingly US centric but that I think is just the website demographic. Actually I think I prefer non-tech articles.
- Technical things that have nothing to do with how amazing or bad Rust, Golang, or a new js fotm is. Hype distortion is bad.<p>- Math stuff<p>- Bug fixes and security things<p>- Career related things<p>I try to avoid flamewar topics and articles that debate gender/race/politics/psychology for the 309838th time. They usually just make me angry, everyone has the same opinion coming out of it as they did going in and when I'm done reading I've learned nothing new. I wish they would be posted less.
I find it amusing that my approach is almost the exact inverse of the OP.<p>- If it has the word "quantum" in it, I almost always click it. I like things that are too technical for me. I really enjoy challenging scientific/ mathmatical articles, they are fascinating & humbling at the same time.<p>- certainly click on very specific programming language links, especially languages that I don't use. I am curious, and try to expose myself to a wide range of languages.<p>- won't read anything from certain domains, anything of questionable quality or click-baity/ tabloidy, eg. buzzfeed.com. Also, go on rants in comments about said questionable websites & how mad it makes me we are driving traffic & ad revenue to those jerks :p<p>- will click on the latest iteration of "MLS tutorial for beginners". I AM the right audience, and am highly interested in the applications of training models.<p>Basically I use HN to learn about things I don't know. I try to expand my horizons, and glean any useful knowledge or information I can from people way smarter and experienced than me.
Inversely correlated to number of comments.<p>Fluff article on bike shedding? 500 comments by know it all.<p>Technical deep dive into exciting new area? No comments.
I rarely click on anything not on the front page from a domain I don’t recognize. Even with all the noscript and ad blockers in the world there are too many baddies out there to justify the risk.<p>As far as the front page goes I lean toward titles that are about (in no particular order):<p>Infosec<p>Politics<p>Databases<p>Kernel / OS<p>News about major tech corps<p>I upvote stories that I want to see a discussion about or stories that I want to have a discussion about.
I look for keywords regarding the tools I use and that describe myself. For me, it's C, C++, random Javascript libraries, databases, certain computer hardware, companies I use, and certain startup business terms. It can be a keyword soup that doesn't make sense and I'll still click on it.<p>I also click on words that appear to be made up, because I've been trained that I'll usually get to read about a cool new library, programming language, or programming tool.<p>In your example, "quantum" is too vague. Even though I've worked for years in quantum-whatever areas, I feel that the word "quantum" doesn't describe myself due to being too general, so it doesn't trigger my attention.<p>So basically, I just scan the front page for words that are attractive to me, never fully reading headlines until I've already decided I'll click on it.
I tend to be a "comments first to see if the source may be interesting" type; this has some subtle effects on my heuristics.<p>1) I tend to avoid (the more obvious) clickbait. Useless information aside it usually spares me the inevitable thread about it being clickbait.<p>2) I tend to avoid political links. The source is usually politically bent, the comments are polarizing, and it's often unnecessarily toxic.<p>3) I tend to avoid failure/scandal articles. They are usually just people acting smug because it was obvious that reboxing pizzas at cost was not a sustainable business model. No substance.<p>4) I tend to enjoy reading technical articles on languages/tools that I use or aspire to use. They're often full of very interesting discussions. I especially like release candidates/patch notes as people often have uses for new features that I hadn't quite considered or groked which makes the value-add all the greater for me.<p>5) I like a lot of articles about potential futures; where we're going with DNA, with space, with physics, etc. These are things well at the periphery of my own knowledge but seeing passionate people talk about what the future could hold excites me for what I have to look forward to in my life.<p>6) Beyond those heuristics I also often do a lot of thread-collapsing when I see things start to devolve into zealous/crude mentality. That being said, I often do follow witty retorts and quips, so there's that.<p>7) I also check out a lot of new languages and tools, not particularly because I want to use them but because I want to know the kinds of problems they perceive with existing tools and how they went about determining solutions to them. It's fun to see some of them start to mature and turn into real things, though (like Rust, which I do explore).<p>As far as <i>actually clicking on the link</i> goes, I usually click on anything that has a few decent comments (or, occasionally, very few comments) and will often read it if it seems highly interesting or opens well and is long-form.
I always click on anything on the front page where I don’t understand the headline. If it’s good enough to get votes _and_ I don’t know what it is, then it’s likely to be a good opportunity for me to learn something new.
If its anything I expect their will be fights and complains. Hell yeah inam clicking those l. For now my mind looks for<p>1. Uber screwed up in a new way.<p>2. How interview process sucks.<p>3. Why this X language / framework sucks or we are leaving this.
I try not to click on nytimes, washingtonpost and all the corporate/state propaganda junk. And most importantly anything with a question for a headline ( except for Ask HN obviously).<p>Other than that, I'm open for anything CS/programming/tech/internet related topics.<p>I hate seeing so much anti-facebook, pro-vegan, climate change, etc obvious agenda pushing by organizations here. Though the mods do a good job of filtering much of it out, but it's obvious there are organizations dedicated to spamming HN to push their agenda.
If it’s trending on the front page and I have no idea what it’s even remotely about, then I’m definitely clicking. Why else am I here if not to be exposed to things outside of my day to day?
1. Will click on anything with the number of comments more than 100.
2. Then will read the first couple of comments
3. Then will decide whether to read the original article
For the main page:<p>1) Anything which has high number of votes and moderate amount of comments is a good technical post.<p>2) Anything with high number of votes and more or less equal (or higher) amount of comments is a controversial post.<p>3) Anything with very high votes and a few comments is a very niche tech post I generally keep an eye on.<p>For everything else,
I just keep an eye on my favurite topics: Emulation, Systems Level Programming, Gamedev<p>There are some posts which do not cross 100 Mark, but still seem to be gold. Trust your favourites
Security vulnerability articles are often good. You get an in depth look at a piece of technology and problem solving related to it.<p>Announcements of company shutdowns can be interesting. The posts themselves tend not to be informative, but the comments can be quite insightful. There are almost always customers who chime in.<p>I'd fully ignore hype of the moment topics like cryptocurrencies or AI. Good articles on these subjects exist, but they are drowned out by noise.
I love reading anything related to JavaScript(ideally React) and cryptocurrencies. I am probably the worst kind of techie in the eyes of quite a few people here.
i think i used to have some inherent heuristics around avoiding content that i assumed would be a little too technical etc, but at some point i kinda dropped all preconceptions and decided to click on random stuff and exit within ~30 seconds if it seemed boring -- i think it's exposed me to a lot of content that has actually been super interesting that i'd otherwise have avoided for fear of not understanding much.
I thought I was mainly looking at two numbers: points and comments. However, after I've built a scatter plot of HN, I realized that's not enough. I still visit HN frequently to look for links with interesting keywords, or from specific domains.<p>But one thing is sure, I seldom look at the user, whether it's green or not.
I’m a non-technical founder, so I click mostly on:<p>Accessibility-related articles (my field)<p>Education/edtech articles (my field)<p>Most Show HNs that I’m in the target demographic for<p>Political articles that have lots of upvotes or comments (though I won’t always have the patience for diving into the comments)<p>Science videos or novelties that appear not to require background knowledge
1. Major events and tech news. I don’t read or listen to many news sites so it helps keep me up on the world.<p>2. New programming languages, especially the hobbyist ones. I like to see what other people think is useful in a language.<p>3. Mainstream psychology articles. I find stuff like biases and decision making fascinating.
I almost never click on a link where the title is the same as the domain name, e.g., "The Story of Foo" where the link is "thestoryoffoo.com"<p>Such links seem to me to be usually publicity-seeking rather than information-providing.
Mostly I click on things that teach me something about programming, things that make me feel like my comment will add to the noise, or things that instinctively make me angry.<p>I avoid everything about startup entrepreneurship like the plague.
I don't actually ever click on links from the homepage— I usually look at the comments first, then apply some heuristics like the types of discussions that are being had to see if I might like the article.
I actively avoid anything to do with geopolitics/US politics/social justice wars/fads (cryptocurrency), and as a result, I find HN (still) useful and informative.
If the title fits one line and I understand every single word from it and can guess what it is about I click. And believe me, it doesn't happen all the time
The science articles are usually pretty interesting. Also, it's a nice place to catch up on tech announcements from Google, Apple, Facebook etc. In fact I'd be happy if these were the vast majority of news items.<p>I tend to avoid links to NYTimes and other paywalled articles. Could be interesting material but usually all I see is "You have read 5/5 articles. Please subscribe or wait until next month."