news.ycombinator.com
Why are there an increasing number of outright unhinged high karma users on HN?https://news.ycombinator.com/item?id=48485429I’ve been noticing this disturbing trend for quite a while. By high karma I mean well over 1000 karma.And some of these leave comments so bizarre, that I can’t find any charitable explanation.Other than perhaps they may be skipping reading every other word or something… and then respond to an imaginary version of the parent.Uncharitably, they could be semi-random keyboard bashing, deranged bashing, emotional tantrums, and so on. Or LLM written… except the style doesnt match.Of course this has always been somewhat prevalent for low karma users, but it’s now creeping up higher and higher.For example, I had a recent bizarre encounter.I had double checked the first 10 pages of google results, in regards to a claim that I implied prior in the comment chain, mentioning that it seemed supportive of the claim.Maybe the novelty of someone doing that, and not just randomly opining, triggered something.Realistically, it didn’t take more than 5 minutes of effort, but still… it would have been 1 toJun 11, 2026 2:07 AM
science.nasa.gov
Jim Irons, Former Landsat Project Scientist, Wins Pecora Awardhttps://science.nasa.gov/missions/landsat/jim-irons-former-landsat-project-scientist-wins-pecora-award/Landsat’s Jim Irons won the prestigious William T. Pecora Award. Irons, now an emeritus scientist at NASA Goddard Space Flight Center, played an integral role in shaping the Landsat program into what it is today.Jun 10, 2026 2:00 PM
news.ycombinator.com
Ask HN: Are we going to see a portable S3 Tables alternative outside AWS?https://news.ycombinator.com/item?id=48475779AWS announced S3 Tables earlier this year, and the integration with Iceberg is clearly the direction for AI/analytics data stacks. But the native S3 Tables implementation is locked inside AWS.For those of us running self-hosted or multi-cloud, what are the current options?Using regular S3 + separate catalog (Hive/Glue/Polaris) – metadata overhead and extra opsApache Ozone / Ceph + Iceberg – seems heavy for smaller teamsSomething else I'm missing?I've seen rumors about open-source projects adding native Iceberg table support directly in the storage layer (basically S3 Tables but portable). Would that actually help? What's the #1 pain point you'd want solved?Curious to hear from people who tried to run Iceberg workloads outside AWS.Jun 10, 2026 1:10 PM
bitemap.food
Show HN: Bitemap – See where everyone would bite a sandwichhttps://www.bitemap.food/sandwich/mcdonalds-cheeseburger-3862dc92I know the internet likes to celebrate perfect bites and debate where to take the next bite of half-finished sandwiches. So I built Bitemap. You see a photo of a sandwich, you tap where you'd take your next bite, and it shows a heatmap of where everyone else bit.It's early. Most of the sandwiches are ones I uploaded to seed it and the maps are still filling in, so some are thin. You can add your own too.No signup, nothing to install. Curious what you think, and where you'd bite. Happy to answer anything about how it's built.Jun 9, 2026 1:29 PM
welter.fuglede.dk
Show HN: Welter's game, a little Nim-like combinatorial gamehttps://welter.fuglede.dkHi HN! Here's a little game I (and my buddy Claude, cf. the obvious CSS gradient tell) set up to explain to my kids some of things we were working on at work.So in the game you're taking turns moving coins down a line until no moves are possible. If you've taken a course on combinatorial game theory, you'll have come across the game of nim and you can think of this one as a variant of nim where there can only be a single coin in every square. The theory of how to play it isn't quite as well-established as for nim though; play with enough coins and I certainly find it quite tricky to mentally reason about whether a position is good or not.One fun fact about Welter's game is that its winning positions form the codewords of a particular error-correcting code called the constant weight binary lexicode (Conway and Sloane, 1986). What that means is that being good at the game is the same as being good at decoding messages in that code. Probably more fun to just play the game though.Jun 9, 2026 11:58 AM
news.ycombinator.com
Ask HN: How do you cope when your startup contracts?https://news.ycombinator.com/item?id=48453224The same general situation has happened to me twice now and I am wondering if it’s something I can break free from or if it’s just the nature of the Startup beast - or what. There seems to be some kind of bubble that starts drying up investment in a startup where I am a technical lead. Both times, things seem to be going well and then 2 years in there are rounds of layoffs due to factors outside my/product’s control where the result is the same. I end up as the last tech generalist. It falls to me to write as much of the code as I can, manage (if any are left) engineers in my department, running support basically on my own, owning a large part of the product roadmap, working with the customers on support and implementations, working with integration partners, plus a slowly expanding list of responsibilities as new stuff falls on reduced teams. I pick up new domains pretty quick (finance, insurance, compliance), switch contexts well. I’ve always been the person who just figures it out -Jun 8, 2026 10:28 PM
blockchainednovel.com
Show HN: Startup sci-fi novel that took me 5 years to writehttps://www.blockchainednovel.com/It started after reading Stephen King's "On Writing" where he likened the art of writing as the unearthing of an archeological site after you stumbled upon a unique bone of a story. His advice was to choose a domain you are deeply intimate with. For me, I've been a struggling startup founder for 15 years—enough material to inspire a novel.A 1,000-word writing exercise turned into a complete 125k-word manuscript over the course of a year.In year 1, I learned the sheer joy of unencumbered creative flow and authentic expression. A similar flow I used to get from coding (and more recently vibe coding). What made it effortless was a mindset that I was writing for the sake of it, not with the intention of publishing.After a year of keeping it close to my chest, I decided to show it to a few close friends. They liked some of it, destroyed some of it. Some ultimately encouraged me to publish it.In year 2, I learned about the chasm between writing for myself and writing for an audience. Nerdy sJun 8, 2026 4:29 PM
bing.com
TEACH YOU A LESSON : STREAM IT OR SKIP IT?http://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a89230bf4f64c83b1810fd3917314d1&url=https%3A%2F%2Fdecider.com%2F2026%2F06%2F05%2Fteach-you-a-lesson-netflix-review%2F&c=7698106904041695569&mkt=en-usWere you bullied in school? Ever wish someone would come in and do to the bullies what they did to you? In a new Netflix series, Teach You A Lesson, that is what happens, with a new government-backed ...Jun 5, 2026 2:45 PM
bing.com
TEACH YOU A LESSON : STREAM IT OR SKIP IT?http://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a8905ed98e24ad5b90baaf9cdf6cc13&url=https%3A%2F%2Fdecider.com%2F2026%2F06%2F05%2Fteach-you-a-lesson-netflix-review%2F&c=7698106904041695569&mkt=en-usWere you bullied in school? Ever wish someone would come in and do to the bullies what they did to you? In a new Netflix series, Teach You A Lesson, that is what happens, with a new government-backed ...Jun 5, 2026 2:45 PM
news.ycombinator.com
Launch HN: Hyper (YC P26) – Company brain to power agentic developmenthttps://news.ycombinator.com/item?id=48387095Hey HN, we’re Shalin & Kanyes, best friends who've been hacking together for 10+yrs, and now founders of Hyper (https://heyhyper.ai/). Hyper is a shared “company brain” that plugs into information flowing inside a company to make AI agents and automations better and ultimately save people time.Models have gotten good enough that they can (mostly) take on long-horizon, complex tasks. We believe the bottleneck now is that these smart-enough models often lack information about your company, which is scattered in people's heads, Slack threads, stale docs, and in back-and-forth convos with AI.MCP is useful for getting some info in front of an agent, but there are problems: (1) Once the session dies, so does the insight, so instead of copy-pasting a whole doc each time you're telling the agent to dig through Drive each time - not much of a win; (2) Even when MCP works, what it gathers isn't comprehensive, because people decide things on a whiteboard, brainstorm out loud, post a little in SlaJun 3, 2026 5:39 PM
news.google.com
Disney+ Removes Over A Dozen 20th Century Studios Filmshttps://news.google.com/rss/articles/CBMijgFBVV95cUxNRjBPaWNpZG5TR1lFN2tvOGNnUTFRT0djdVVDSmtCVzZFMTRWTEF4aTk1b2g4WHRwR1lLbnE0MjZFdnlKanQtRDVnSnJhTHhtUVZ0QmR4Q0xabDVkSDhYWkdKS2pnSlluYlE2TXRpcy1tNW53Mnh3SFB4QzBvNlZiMnZzdXV0UTE2cHZUQ21B?oc=5Disney+ Removes Over A Dozen 20th Century Studios Films whatsondisneyplus.comJun 1, 2026 7:00 AM
pictolab.io
Show HN: Pictolab (online HDR image editor)https://pictolab.io/Source code: http://github.com/anchpop/pictolabI love HDR, but until recently dynamic HDR content was hard to render in the browser. But now, with browsers shipping support for WebGPU and canvases with extended color spaces, it's actually feasible.So what I wanted to was make an image editor capable of doing the simple edits I often need to do, but with full HDR support, which is rare in browser editors. Turning up the brightness on an HDR-capable monitor is a great way to see what your monitor can do :PI tried to go overboard in every way I could. So brightness/saturation/hue adjustments use the okLCH color space, and there's an option to do a "content-aware resize" with seam carving, which is fast because it's parallelized using Sam Westrick's triangle-blocking algorithm [0].Also, the site supports importing HEIC photos as taken with an iPhone, which is surprisingly hard to do correctly. Libraries like libheic don't fully support the format used by iPhones, so if used naively the HDRMay 31, 2026 9:45 PM
vibebnb.fyi
Show HN: I brought back Airbnb categorieshttps://www.vibebnb.fyi/I miss airbnb categories. It was fun and I loved being able to just browse thorugh cool listings. I'm not exactly sure why they removed it. Then, a couple weeks ago, they hit me with some A/B test and I had access to a "what" freeform textbox to custom filter results. Now apparently they removed that too.At this point, I wasn't happy - so I decided to look into how to bring it back. I noticed they still use some of the categories pages for SEO. For example: https://www.airbnb.com/united-states/stays/ski-in-ski-outThat gave me hope - those filters still existed somewhere. Turns out they were only removed from the UI and the search backend still supports all of them.So after a bit of googling, I figured out how it works. Airbnb's search runs on a GraphQL endpoint you can hit anonymously, and the categories map to params on the search URL that I recovered and verified one by one:- Property types: property_type_id[]= (Chalet=22, Cabin=4, Treehouse=6, Castle=5), still filtering even though May 29, 2026 7:53 PM
news.ycombinator.com
Show HN: Sixbpm – a free thing that slows your breathing downhttps://news.ycombinator.com/item?id=48320579i made https://sixbpm.com with the help of an llm, im not an engineer by trade just someone whos suffered from burnout from cluster headaches and has struggled recently with rebound headaches.there is nothing to install, no account, just a web page.You lie down, put the phone face-down on your chest and tap start.It listens to your breathing through the accelerometer for a minute, then plays two tones - one for the inhale, one for the exhale - and slowly stretches them out until you're around six breaths a minute. Fifteen minutes total.Six breaths a minute is roughly the resonance frequency of your cardiovascular system. At that rate with a long exhale your heart rate variability, blood pressure, and breathing sync up and your vagal tone climbs. There's real research behind this going back to Lehrer and Vaschillo in the early 2000s. It's also what a $320 FDA-cleared device called RESPeRATE does and the patent expired, the protocol is in the public domain, it's two tones and a timer.MosMay 29, 2026 8:29 AM
news.ycombinator.com
How to prevent ideological capture of sites like Wikipedia, HNhttps://news.ycombinator.com/item?id=48290346Someone submitted a link [1] to an article on Wikipedia's whitewashing of Mao titled Wikipedia Whitewashes Mao – The Anatomy of Ideological Capture [2], a topic which in a more generalized form is of interest to the current average HN denizen given the prevalence of Wikipedia in LLM training data. It did not take long for the flaggers and down-voters to take down the article which has only managed to get a few comments, the top one claiming the article to be 'a pretty political take on a Wikipedia article'. I think the subject deserves more discussion since it points out a clear problem with sources like Wikipedia but also because of the way this article was quickly flagged to death. The flagging and down-voting mechanisms make sites like HN vulnerable to the same type of ideological capture as what is discussed in the linked article in that a relatively small group of motivated users can suppress 'unwanted' subjects while boosting those which fit some desired narrative.[1] https://newMay 27, 2026 6:14 AM
github.com
Show HN: Monkdev is a toolkit and methodology for coding with LLMshttps://github.com/oeo/monkdevI'm sure many people have some solutions similar to this but I've been using some variation of this since roughly the release of Opus 3 to get higher quality results. Like, significantly higher. So I decided to make it into a more structured package.Problem: bad decisions and short sighted solutions due to the LLM not having enough context before acting. They'd read one file, or a few segments of a set of files using a super conservative offset and limit, guess about the rest, apply a patch, run a unit test and then tell you with its entire chest that the code is production-ready.Those bandaids over time became debt, drift, complexity. Unnecessary LOC in a codebase where I find myself wondering is this thing even making me more efficient or should I be in VIM more often and just save myself the API costs (I use OpenCode + OpenRouter and have been known to blast through hundreds of dollars a day in spend using Frontier models).I found that, for whatever reason, LLMs respond to role-playMay 26, 2026 7:08 PM
news.ycombinator.com
Ask HN: Encouraging a child's gaming PC build despite fear of gaming addiction?https://news.ycombinator.com/item?id=4826332112yo son is doing dedicated research on gaming PC parts and binge-watches videos on how to build (he's not a native English speaker and has no real electronics skills, so this is quite impressive tbh). Me the dad, however, fears that I'm witnessing a FPS addict being born.What's a good Middle Way here? Kid actually building the machine all by himself -- 100% yes, even if it's quite expensive; kid then Counter Striking, Call of Dutying etc for hours on dat 27" screen -- ugh!What complicates things is that I myself am more of a "computational minimalist" using 15 year old machines in a cli-only environment etc. I've tried to show him That Path also ("dude, here's some Turbo Pascal for ya!"), but, well, obviously he's preferring those addictively engineered and GPU-hungry multiplayer games with his friends any day.He's a good kid all in all, actively engaged in competitive rowing, with good grades etc. Plays chess also (but all the other folks are FPSing or Minecrafting, so there's that).May 25, 2026 4:11 AM
news.ycombinator.com
Tell HN: The Threat to US Citizen's ID / Voting Is Private Serviceshttps://news.ycombinator.com/item?id=48238426THE REAL RISK AROUND VOTER ID NOW COMES FROM PRIVATE ENTERPRISEThis post has been condensed due to character limitations on HN posts.[...]What's happening is that private entities are curating the databases and operating automated adjudication services which the government contracts to utilize. "But it's little things, administrative things." Yeah maybe, but look where they are in the kill chain. During approximately the week surrounding 15-May-2026 I was denied the "privilege" of scheduling an appointment at the United States Post Office by one of these "administrative things" and there is no particular reason to think that the government runs this service themselves, relying entirely on compute and data which they control. So THIS is how the woodpecker destroys civilization; or at least our representative democracy. There is no appeal. There is no overt denial at the time of service either.Here's how it works in real life:The United States federal government operates Login.Gov, a sinMay 22, 2026 4:54 PM
classkeep.app
Show HN: ClassKeep – Booking, credits, and smart waitlists for boutique studioshttps://www.classkeep.appHi HN — I built ClassKeep.The trigger: every small studio owner I know (pilates, yoga, pottery, even a dog trainer) runs the operational side of their business on WhatsApp + a Google Sheet, or on one of the expensive incumbents (Mindbody/Glofox) they actively dislike. The single biggest leak is last-minute cancellations: someone bails at 6am and the seat just goes empty, because there's no fast way to notify a waitlist.ClassKeep tries to remove that end-to-end:- Public booking page on a free *.classkeep.app slug (custom domain optional), with a drag-and-drop page builder so it doesn't look like a generic SaaS template. - Credits, drop-ins, and recurring subscription plans, all via Stripe Connect — the studio gets paid directly, I never touch the money. - A waitlist that auto-notifies via email + web push + SMS the instant a seat opens; first tap wins. This is the part most studios told me they'd pay for on its own. - 1-tap tablet check-in for instructors so they're not fumbling with a May 19, 2026 4:38 PM
github.com
Show HN: Forge – Guardrails take an 8B model from 53% to 99% on agentic taskshttps://github.com/antoinezambelli/forgeHi HN, I'm Antoine Zambelli, AI Director at Texas Instruments.I built Forge, an open-source reliability layer for self-hosted LLM tool-calling.What it does:- Adds domain-and-tool-agnostic guardrails (retry nudges, step enforcement, error recovery, VRAM-aware context management) to local models running on consumer hardware- Takes an 8B model from ~53% to ~99% on multi-step agentic workflows without changing the model - just the system around it- Ships with an eval harness and interactive dashboard so you can reproduce every numberI wanted to run a handful of always-on agentic systems for my portfolio, didn't want to pay cloud frontier costs, and immediately hit the compounding math problem on local models. 90% per-step accuracy sounds great, but with a 5-step workflow that's a 40% failure rate. No existing framework seemed to address this mechanical reliability issue - they all seemed tailor-made for cloud frontier.Demo video: https://youtu.be/MzRgJoJAXGc (side-by-side: same model, sameMay 19, 2026 12:23 PM
