Show HN: Nekoweb – a retro static web hosting https://ift.tt/ZXFQ0cm

Show HN: Nekoweb – a retro static web hosting https://nekoweb.org/ February 26, 2024 at 02:33AM

Show HN: Continuous-eval – Granular evaluation of GenAI pipelines https://ift.tt/PtCLhxn

Show HN: Continuous-eval – Granular evaluation of GenAI pipelines Hi HN - we are the creators of “continuous-eval”, an open-source tool to test and evaluate generative AI apps. "Continuous-eval" came from our efforts to measure, validate and improve the reliability of a finance AI copilot we were developing for banks. End-to-end evaluation was not enough for us. We wanted to have granular evaluations that help pinpoint the bottlenecks and identify what / how to improve. We’ve since developed more metrics and made the framework more flexible so it can evaluate components like agent tool use, code change, retrieval steps, etc. Let us know what you think of our approach to GenAI App evaluation. https://ift.tt/oPqb51D February 26, 2024 at 12:11AM

Show HN: Task manager with bear notes style tagging system https://ift.tt/m9vDo8C

Show HN: Task manager with bear notes style tagging system https://hyperaide.com/ February 25, 2024 at 02:27AM

Show HN: Psfiles – a CLI tool to monitor file system activity of a Linux process https://ift.tt/tDfwlhT

Show HN: Psfiles – a CLI tool to monitor file system activity of a Linux process Psfiles is a simple utility to view file system activity of Linux processes. Features: - start new process or attach to existing one and trace its file system activity, - output results to standard output or save results to file, - custom results sorting and filtering. https://ift.tt/9Er8psU February 25, 2024 at 01:08AM

Show HN: Logo Generation with AI https://ift.tt/4VeIN3w

Show HN: Logo Generation with AI https://mylogo.ai February 25, 2024 at 12:33AM

Show HN: I built jq-like scriptable tool to query CSV and JSON with SQLite https://ift.tt/7WXNc3j

Show HN: I built jq-like scriptable tool to query CSV and JSON with SQLite https://ift.tt/CmIvjyt February 24, 2024 at 11:49PM

Show HN: Consol3 – A 3D engine for the terminal that executes on the CPU https://ift.tt/zkohElB

Show HN: Consol3 – A 3D engine for the terminal that executes on the CPU Hi all This has been my hobby project for quite a few years now It started as a small engine to serve as a sandbox to try out new 3d graphics ideas After adding many features through out the years and re-writing the entire engine a few times, this is the latest state It currently supports loading models with animations, textures, lights, shadow maps, normal maps, and some other goodies I've also recently added voxel raymarching as an alternative renderer, along with a fun physics simulation :) https://ift.tt/MFR7Qw9 February 24, 2024 at 07:47AM

Show HN: Refractify: optical software against Myopia https://ift.tt/S1Vtba0

Show HN: Refractify: optical software against Myopia Last summer there was an Ask HN[1] about a Nature article that said bluring the blue and green color channels on screen may be good against early myopia development. The OP wanted such software and there was none available. So I quit my job and implemented this software, did a short video with a 3D artist about it. Turns out marketing is expensive, so I made an open source browser extension version too. How it works? There is a small neural network on the retina that tries to detect if the eye is far-sighted(most people are born far-sighted), and it is producing dopamine to slow or increase eye growth rate. It is not very smart, and if you do a lot of near-work it can think you are still hyperopic, causing further myopia progression. So, based on the refractive properties of the eye the software calculates the signal that would convince the retinal neural network that the eye is long enough, so it would produce dopamine, a known signal to stop axial eye growth. (based on myopic defocus LCA from the papers[2][3]) Some myopia control techniques work similarly, like MiSight and Hoya lenses. Since then I got a Neurobiologist co-founder and the goal is to best understand the Retinal NN to create the best anti-myopic effect that does not interfere with productivity. The effect can be tried live on the site. Also check out the github repo. Any questions suggestions welcome! [1] https://ift.tt/Ypie0zZ [2] https://ift.tt/E4doyHO [3] https://ift.tt/mlM6rgv... https://refractify.io/ February 24, 2024 at 12:17AM

Show HN: Babel – browse every book that will ever be written using simple math https://ift.tt/ubhp0rA

Show HN: Babel – browse every book that will ever be written using simple math https://ift.tt/kUIzawX February 23, 2024 at 10:37PM

Show HN: I made a multiplayer browser game https://ift.tt/4kRA5bq

Show HN: I made a multiplayer browser game http://boxfight.xyz/ February 23, 2024 at 01:31AM

Show HN: Learn Game Theory Optimal Poker Preflop with Spaced-Repetition https://ift.tt/ov5S12L

Show HN: Learn Game Theory Optimal Poker Preflop with Spaced-Repetition Hi HN, Sharing my poker preflop trainer product, a subset of my Live Poker Theory project. https://ift.tt/BQczFwm Live Poker Theory helps translate complex poker solver strategy to actionable strategies while playing and aims to make studying poker more efficient and more fun. While I usually try to focus on sharing it in poker communities, I saw a few poker articles frontpage this site so I figure it doesn't hurt to share it here. In case you don't know, before 2015 most poker software could only calculate "all-in equity" - if there was no game tree and players could only go all-in or fold. These days, solvers can calculate the full game tree, with a lot of assumptions, and we can use them to generate preflop charts. Sometimes people call this GTO (game-theory optimal) though I prefer the term "theory-based" to recognize how frequently you want to diverge from equilibrium even if you've studied it. Preflop refers to the first two cards you're dealt and the first round of betting which is a very fundamental street. Preflop is a good example of where I find it useful to study equilbrium even if you might diverge in practice - for example, once you understand how often a player should be 3-betting you (re-raising you after you've raised), if you know someone doesn't do it with hands like Ace-Five suited, you can fold hands you'd otherwise continue with. But it's helpful to understand a strong player should frequently be 3-betting Ace-Five suited. Some other info that might be helpful: 1) Rake refers to whether the "house" removes money from the pot which happens at most low-stakes games. Higher stakes games tend to be "time" games where the players simply pay an hourly fee so there's no effect of rake on the game itself. That also may be true at an unraked home game. 2) A straddle is an optional third blind that's often strongly encouraged as something everyone at the table does, and of course the charts are different with that third blind Spaced-repetition is something that only my trainer does, while it's a well understood concept on places like Hacker News, it's not well understood by the poker community. Even if you plan to make adjustments against certain players, there's good reasons to memorize a preflop chart. It helps you stay disciplined if you're "tilted" if you know what a reasonable baseline strategy is. It also helps you clearly define your postflop strategy, both while doing solver work and while playing. For examlpe, frequently the best river bluffs are the "bottom of our range", since our worst hands beneift the most from our opponent folding. But the "bottom of our range" is only clearly defined if our range is clearly defined, so if you've memorized your preflop range, you'll have a better understanding of your overall strategy postflop. One last important note, the charts are based on a 2.5x raise, so in a 2/5 game, a raise to $12, which is fairly rare to see in practice. If you have the solver raise 3x or 4x, the overall strategy is much much tighter. While this is more correct against perfect opponents, in practice frequently we're against weaker opponents and we'd rather play a looser range since we'll have an advantage postflop. By studying 2.5x, we keep a more reasonable loose range but still let the solver give us a reasonable baseline of hands to play. Currently, you can try out Tournament, 50 Big Blind stack depth, Raise First In without an account. With an account and for free, you can study cash or tournament "Raise First In" (whether to raise or fold if it folds to you), "Vs Open" (whether you raise, fold, or call if someone raises before you act), for free, for cash or tournaments. I have paywalled only cash game BB defend (if someone raises when you're in the big blind) for $10/mo or $59/year. I'm also actively adding more preflop charts, developing postflop content with spaced-repetition and a native mobile app. https://ift.tt/BQczFwm February 23, 2024 at 01:06AM

Show HN: I scraped 200M Shopify products to build a search engine https://ift.tt/EBTcx5v

Show HN: I scraped 200M Shopify products to build a search engine Hi HN! In December I launched an MVP for Agora here: https://ift.tt/9NiLSqn After posting, we got thousands of users and hundreds of comments with valuable feedback from the community. I spent a couple sleepless nights frantically pacing around my room trying to keep the product live and, relatively, performant. After getting some sleep, I got back to work to make the product better. A few updates: 1. We've grown from 25 million to 200 million products on Shopify and WooCommerce. The team at WooCommerce reached out after the HN launch to help us figure out how to index their stores. Similar to Shopify, we found that there’s a public file available for all stores that use Wordpress and WooCommerce at [Base URL]/wp-json/wc/v1/products. For example, the file for Good Works Tractors is available here: https://ift.tt/zXDjAxg... So I bought a list of 3.5 million active WooCommerce stores on a website called BuiltWith, adapted the product data model, and started the crawler to go down the list. We've indexed around 515k stores so far. 2. We improved the search experience. We're using Mongo to host the 200 million product records. First, we switched from Mongo Atlas Search to Typesense. After testing Typesense with our product records, we found most searches to be under 200ms. We're not storing the product images which slows down the loading speed at times. This week, we set up a server using Paperspace to run SBERT embeddings on a GPU (new to the AI workflow so apologies if I get the lingo wrong). We quickly realized that the dimension size of the embeddings matters a lot here, given the size of the data set. The GPU is still running to process all 200 million records and we're about a week away from releasing AI-powered search. 3. We localized the user experience. There's now frontend and backend IP detection to only show users products that are 'based in' or 'ship to' their specific country. This 'ships to' filter (i.e. stored in all Shopify stores in the /meta.json route like https://ift.tt/psmLtSC ) significantly slows down the search results but we're trying to get creative on the loading process and animation. For example, we're using Revalidating on Next.JS to give several pages a 'hard coded' feel and the data refreshes every 60 seconds. https://ift.tt/ZchNm9G... 4. We got our first few paying customers. Store owners can sign up for free to track their store's performance on Agora. We validate that they are the store owner by making sure the email address and store URL match on sign up, and then send them an email verification link. They can upgrade to a subscription tier to 'verify' their products to get better placement in relevant search results. Additionally, they can pay to 'boost' products and guarantee that they'll show up in the first row of results. Given the high purchase-intent searches on Agora, I'm finding this to be the right business model. The next challenge to solve: We need to improve the quality of products on Agora. There's a lot of resellers, dropshipping stores, and low quality images. Now, just because a product is sold on a reseller or dropshipping website, doesn't mean it's a bad product. There's a lot of exceptions and edge cases to solve. One potential solution: we're considering coming up with an "Agora Score" that takes in several factors including the image quality, store name, brand name, website SEO, etc. to tell users how trustworthy we think the product is. I'd love any feedback or advice. I did solve my original problem of finding 'red shoes' for my wife, but inadvertently created more problems for myself. I'm loving every minute of it though. My wife jokes that everything is now "Agora this...Agora that". Open to any advice on that as well. https://ift.tt/S5qNZj7 February 22, 2024 at 04:04AM

Show HN: Wiwo – Find events happening around you. Google Maps, but for events. https://ift.tt/xEwY8ab

Show HN: Wiwo – Find events happening around you. Google Maps, but for events. Hey everyone in the USA! I'm developing a map-based event discovery platform that helps people find unique and exciting events in their city or area. I imagine that it could help young locals who are searching what is to be attended in their area & tourists who are visiting and want to find what to do in the area. But before I unleash it on the world, I need your help! One major pain point that I am trying to solve is information overload and difficulty in discovering relevant events nearby. Traditionally, people might rely on social media, local newspapers, or word-of-mouth to find events happening around them. These methods can be: Scattered: Information is spread across different sources, making it time-consuming to search. Unfiltered: Showing a large amount of events, not necessarily relevant to individual interests. Limited information: Often lacking key details like event descriptions, schedules, or ticket options. So my MVP address these issues by: Aggregating: Bringing together event information from multiple sources in one place. Personalization: Using filters and user preferences to recommend relevant events. Visualization: Displaying events on a map, allowing users to discover events based on location and proximity. Detailed information: Providing comprehensive descriptions and links to purchase tickets. What other features I plan to build: Build-in resell marketplace, own event upload, live-streaming, PPV and more. However I find it difficult to get any feedback on the platform. I've tried paid ads, posting to other Early Adopters websites, personally reaching out to people, who I think are target users, but do not get any meaningful feedback. So I'm eager to hear your honest opinions: What works well in the platform? What needs improvement? Are there any features you'd like to see added? Your feedback will be invaluable in shaping the future of Wiwo! P.S. For now only US users are needed, as other won't find events in their areas :) https://ift.tt/vkHtcfr February 22, 2024 at 02:28AM

Show HN: An Experiment with One-Feature Tool Made $7164/Mo https://ift.tt/gEPh8Zv

Show HN: An Experiment with One-Feature Tool Made $7164/Mo My Raw Story on coming up with an idea, building and growing it. It's very detailed, with the purpose of giving another founder an insider look at one way of doing it. In January I launched an indexing tool called Index Rusher, that forces google to index your pages quicker, to get ranked for SEO faster. This whole project was something I needed myself since I got over 20 products and paying for an external one would simply cost too much. My initial idea was that I would just build an internal tool for my use, that has only 1 feature. No UI really, just 1 button. In the middle of the process, I realized that I could actually run an experiment and launch this tool publically with just one feature. Super simple. I hired a dev who spent a month building it. It looked super easy at first, but it turned out there were so many hidden snakes on the way. Troubles with sitemaps, google APIs, and more. 1 month later I launched it (In Jan). The launch didn't go so great, but I didn't really have high hopes. Because nobody knew about this tool, I had no traffic on the site. I still sold several licenses, which made me pretty happy, it felt like validation, people needed it, even if it solved such a narrow problem. At that point, I declared my next stage of the experiment: Growing the traffic and revenue. I've done a number of growth hacks in the next 30 days, resulting in over $7k in revenue, but what's more important, the traffic on the site has grown a lot and stays high and growing. This means I've done a pretty good job on organic growth too, which will just accelerate over time. Here is what I've done: Cross-linking. I added links in the footer on my other products. This is one hidden effect of having multiple products. Each may serve as a lead magnet for the other one. In my case, I have the same audience for all my tools, people who love one of my tools often check out the rest. Being visible on social media. I monitor discussions around the Google Indexing topics and add my replies there. I don't just spam in replies with my tools, in most cases, I genuinely answer and bring value. If my reply gets a reply, I may include my URL in the next reply. Social Media and Blog posts. I've posted several posts about Growth, where I mentioned Indexrusher since I actually use it for me Growth. Traffic from Directories. This one was the top channel of growth. Over 50% of the paying users arrive from web directories. I used a tool that listed Index Rusher on 100 directories & websites. Sponsored listings. I "sponsored" directories to place a banner for my tool on the top of their page/list. Seeing the effect of "boosted" listings. The ROI was good. About $2.5k of revenue came in from these boosts. Affiliate partners Made a deal with a few affiliate partners who reached out to me on X and he drove a decent amount of traffic and paid users to me since he was launching on PH the same week, The total economy of the project now Dev costs: $1500*3=$4500 - Godaddy domain: $9 - hetzner Hosting: $10/mo - landing page on Unicorn Platform: $9/mo - cost of sponsorships: $800 - Affiliate payouts: $150 - listingbott for backlinks: $499 - seobot ai for blog: $99 - Stripe fees: $654 Total cost: $6711 Revenue: $7164 Profit: $453. So, it's profitable! My next steps will be 1) Promote it to 100,000+ users of my Website Builder and reach out to more website builders and pitch them the integration 2) Increase Word-of-mouth effect 3) Perhaps try some paid ads 4) Add automated emails to remind about Index Rusher users who signed up but didn't buy 5) Launch a directory as a lead magnet 6) Launch little free tools as lead magnets 7) Product Hunt launch 8) AppSumo launch I will make a new post in a month describing how it went. February 22, 2024 at 02:13AM

Show HN: jSuites v4 - A library of ultra-light components and plugins free (MIT) https://ift.tt/dhvxQIo

Show HN: jSuites v4 - A library of ultra-light components and plugins free (MIT) https://ift.tt/u6fDKsL February 21, 2024 at 10:47PM

Show HN: LoraLand – 25 fine-tuned LLMs that beat GPT-4 https://ift.tt/JrxoWmb

Show HN: LoraLand – 25 fine-tuned LLMs that beat GPT-4 Hi all, today we're excited to launch LoraLand: 25 fine-tuned mistral-7b models that outperform #gpt4 on task-specific applications ranging from sentiment detection to question answering. All 25 fine-tuned models… - Outperform GPT-4, GPT-3.5-turbo, and mistral-7b-instruct for specific tasks - Are cost-effectively served from a single GPU through LoRAX - Were trained for less than $8 each on average You can prompt all of the fine-tuned models today and compare their results to mistral-7b-instruct in real time! We'd love to hear comments and feedback from the community https://ift.tt/0pXFhgf February 20, 2024 at 10:18PM

Show HN: Hyperdiv – Reactive, immediate-mode web UI framework for Python https://ift.tt/RFA9rTq

Show HN: Hyperdiv – Reactive, immediate-mode web UI framework for Python Hello HN, I'm releasing Hyperdiv ( https://hyperdiv.io ), a framework for rapidly developing reactive browser UIs in Python, with immediate-mode syntax and using Shoelace ( https://shoelace.style ) as its built-in component system. This short coding video will give you a good idea of what it is: https://www.youtube.com/watch?v=4XJKfxaqvGE I wrote a brief article about the motivation and approach: https://ift.tt/mP3sXFk Hyperdiv doesn't aim to compete with serious full-stack frameworks. The core aim was to make it easy and fast to prototype apps and build UI-based tools. I was originally motivated by internal tools at work -- feeling the need to quickly put together UI-based tools to share with both technical and non-technical coworkers, without having to stand up and maintain a full internal stack. This is my first major open source release. I really appreciate your feedback and support. - Marius https://ift.tt/OHvQmzS February 20, 2024 at 07:53PM

Show HN: DMARC Checker https://ift.tt/uy2wWKR

Show HN: DMARC Checker https://ift.tt/FA7kVjp February 20, 2024 at 10:22PM

Show HN: CaveRibbon (SFCave Remake) https://ift.tt/9xyetF1

Show HN: CaveRibbon (SFCave Remake) https://ift.tt/kpP90FG February 19, 2024 at 05:02PM

Show HN: Tl;Dr Voters – Simplifying Democracy with AI Powered Ballot Summaries https://ift.tt/97Ge5Hs

Show HN: Tl;Dr Voters – Simplifying Democracy with AI Powered Ballot Summaries https://ift.tt/OqWymu6 February 19, 2024 at 10:54PM