Most review tools dump a PR diff into a model and hope it finds bugs. The model sees added/removed lines, hunk headers, context lines. It has no idea that the function it's looking at is called by x other functions across y files, or that a type change here breaks an interface three directories away.
The triage layer parses source code into ASTs using tree-sitters, extracts semantically meaningful entities (functions, classes, methods, structs), and builds a cross-file dependency graph. It ranks every changed entity by transitive blast radius. Cuts the review surface by 80-90%, and increases the attention score on the bug significantly. Now I am sure it can be out of distribution few times but for fast code reviews this tradeoff is worth making.
Once you've narrowed the problem to "here are the n riskiest entities in this PR," you don't need a frontier model. You need a model that just knows your code. A 7B fine-tuned on your codebase knows your patterns, your conventions, your common bugs. Structural triage handles the global reasoning that results in your model handling the judgment call really well.
Commands:
- inspect diff - entity-level diff with risk scoring and blast radius
- inspect predict - show which unchanged entities are at risk of breaking
- inspect review - structural triage + LLM review
- inspect pr - review a GitHub PR
21 language parsers. Written in Rust. Open source.
Github: https://github.com/Ataraxy-Labs/inspect"},"title":{"matchLevel":"none","matchedWords":[],"value":"Structural and semantic component for improving code reviews with local models"}},"_tags":["story","author_rs545837","story_47686246","ask_hn"],"author":"rs545837","created_at":"2026-04-08T06:44:42Z","created_at_i":1775630682,"num_comments":0,"objectID":"47686246","points":2,"story_id":47686246,"story_text":"I was curious in improving code reviews because they still suck, so researching on a triage layer that you can attach to your local LLMs/api calls for better code reviews.
Most review tools dump a PR diff into a model and hope it finds bugs. The model sees added/removed lines, hunk headers, context lines. It has no idea that the function it's looking at is called by x other functions across y files, or that a type change here breaks an interface three directories away.
The triage layer parses source code into ASTs using tree-sitters, extracts semantically meaningful entities (functions, classes, methods, structs), and builds a cross-file dependency graph. It ranks every changed entity by transitive blast radius. Cuts the review surface by 80-90%, and increases the attention score on the bug significantly. Now I am sure it can be out of distribution few times but for fast code reviews this tradeoff is worth making.
Once you've narrowed the problem to "here are the n riskiest entities in this PR," you don't need a frontier model. You need a model that just knows your code. A 7B fine-tuned on your codebase knows your patterns, your conventions, your common bugs. Structural triage handles the global reasoning that results in your model handling the judgment call really well.
Commands:
- inspect diff - entity-level diff with risk scoring and blast radius
- inspect predict - show which unchanged entities are at risk of breaking
- inspect review - structural triage + LLM review
- inspect pr - review a GitHub PR
21 language parsers. Written in Rust. Open source.
Github: https://github.com/Ataraxy-Labs/inspect","title":"Structural and semantic component for improving code reviews with local models","updated_at":"2026-04-08T06:50:12Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"prat"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":" Labs-on-a-chip that you can shrink to fit "},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["8090"],"value":"http://www.newscientist.com/article/mg20427305.800-labsonachip-that-you-can-shrink-to-fit.html?DCMP=OTC-rss&nsref=online-news"}},"_tags":["story","author_prat","story_889557"],"author":"prat","created_at":"2009-10-19T05:40:12Z","created_at_i":1255930812,"num_comments":0,"objectID":"889557","points":2,"story_id":889557,"story_text":"","title":" Labs-on-a-chip that you can shrink to fit ","updated_at":"2023-09-06T20:09:58Z","url":"http://www.newscientist.com/article/mg20427305.800-labsonachip-that-you-can-shrink-to-fit.html?DCMP=OTC-rss&nsref=online-news"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"boredemployee"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"China's New Lab, 7,800 Feet Under a Mountain, to Hunt Dark Matter"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"https://www.businessinsider.com/china-worlds-biggest-deepest-lab-hunt-for-dark-matter-cjpl-2024-1"}},"_tags":["story","author_boredemployee","story_39188555"],"author":"boredemployee","created_at":"2024-01-30T10:46:31Z","created_at_i":1706611591,"num_comments":0,"objectID":"39188555","points":2,"story_id":39188555,"title":"China's New Lab, 7,800 Feet Under a Mountain, to Hunt Dark Matter","updated_at":"2024-09-20T16:19:02Z","url":"https://www.businessinsider.com/china-worlds-biggest-deepest-lab-hunt-for-dark-matter-cjpl-2024-1"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"andygcook"},"title":{"matchLevel":"none","matchedWords":[],"value":"Notion Hits $800M Valuation in New Funding"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://www.theinformation.com/articles/collaboration-app-notion-labs-hits-800-million-valuation-in-new-funding"}},"_tags":["story","author_andygcook","story_20479081"],"author":"andygcook","children":[20479272],"created_at":"2019-07-19T14:47:58Z","created_at_i":1563547678,"num_comments":2,"objectID":"20479081","points":5,"story_id":20479081,"title":"Notion Hits $800M Valuation in New Funding","updated_at":"2024-09-20T04:39:38Z","url":"https://www.theinformation.com/articles/collaboration-app-notion-labs-hits-800-million-valuation-in-new-funding"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"justswim"},"title":{"matchLevel":"none","matchedWords":[],"value":"Notion Valued at 800m"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://www.theinformation.com/articles/collaboration-app-notion-labs-hits-800-million-valuation-in-new-funding"}},"_tags":["story","author_justswim","story_20476091"],"author":"justswim","created_at":"2019-07-19T04:07:34Z","created_at_i":1563509254,"num_comments":0,"objectID":"20476091","points":3,"story_id":20476091,"title":"Notion Valued at 800m","updated_at":"2024-09-20T04:39:21Z","url":"https://www.theinformation.com/articles/collaboration-app-notion-labs-hits-800-million-valuation-in-new-funding"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"dirtyaura"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"Collaboration App Notion Labs Hits $800M Valuation in New Funding"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://www.theinformation.com/articles/collaboration-app-notion-labs-hits-800-million-valuation-in-new-funding?shared=051dd431acd3603d"}},"_tags":["story","author_dirtyaura","story_20483065"],"author":"dirtyaura","children":[20483080],"created_at":"2019-07-19T22:44:44Z","created_at_i":1563576284,"num_comments":1,"objectID":"20483065","points":2,"story_id":20483065,"title":"Collaboration App Notion Labs Hits $800M Valuation in New Funding","updated_at":"2024-09-20T04:40:03Z","url":"https://www.theinformation.com/articles/collaboration-app-notion-labs-hits-800-million-valuation-in-new-funding?shared=051dd431acd3603d"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"urs2102"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"Collaboration App Notion Labs Hits $800M Valuation in New Funding"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://www.theinformation.com/articles/collaboration-app-notion-labs-hits-800-million-valuation-in-new-funding?shared=051dd431acd3603d"}},"_tags":["story","author_urs2102","story_20476010"],"author":"urs2102","created_at":"2019-07-19T03:42:33Z","created_at_i":1563507753,"num_comments":0,"objectID":"20476010","points":2,"story_id":20476010,"title":"Collaboration App Notion Labs Hits $800M Valuation in New Funding","updated_at":"2024-09-20T04:39:21Z","url":"https://www.theinformation.com/articles/collaboration-app-notion-labs-hits-800-million-valuation-in-new-funding?shared=051dd431acd3603d"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"colinprince"},"title":{"matchLevel":"none","matchedWords":[],"value":"What user testing taught me about writing news for young people"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://medium.com/bbc-news-labs/beyond-800-words-what-user-testing-taught-me-about-writing-news-for-young-people-50c31f9f4dd1"}},"_tags":["story","author_colinprince","story_18728815"],"author":"colinprince","created_at":"2018-12-20T20:28:42Z","created_at_i":1545337722,"num_comments":0,"objectID":"18728815","points":2,"story_id":18728815,"title":"What user testing taught me about writing news for young people","updated_at":"2024-09-20T03:28:31Z","url":"https://medium.com/bbc-news-labs/beyond-800-words-what-user-testing-taught-me-about-writing-news-for-young-people-50c31f9f4dd1"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jonas21"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Alphabet's Sidewalk Labs to turn 800 acres of Toronto into an \u201cinternet city\u201d"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"https://www.theverge.com/2017/10/17/16488942/alphabet-sidewalk-labs-toronto-quayside"}},"_tags":["story","author_jonas21","story_15494009"],"author":"jonas21","children":[15494312,15494324,15494332,15494341,15494858,15494967,15495191,15495599,15495990,15496280,15496716],"created_at":"2017-10-17T19:22:50Z","created_at_i":1508268170,"num_comments":93,"objectID":"15494009","points":187,"story_id":15494009,"title":"Alphabet's Sidewalk Labs to turn 800 acres of Toronto into an \u201cinternet city\u201d","updated_at":"2024-09-20T01:34:15Z","url":"https://www.theverge.com/2017/10/17/16488942/alphabet-sidewalk-labs-toronto-quayside"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jackchristopher"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"Lung-on-a-chip could replace countless lab rats"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"http://www.newscientist.com/article/mg20227126.800-lungonachip-could-replace-countless-lab-rats.html"}},"_tags":["story","author_jackchristopher","story_663421"],"author":"jackchristopher","created_at":"2009-06-18T13:41:01Z","created_at_i":1245332461,"num_comments":0,"objectID":"663421","points":2,"story_id":663421,"title":"Lung-on-a-chip could replace countless lab rats","updated_at":"2024-09-19T16:41:27Z","url":"http://www.newscientist.com/article/mg20227126.800-lungonachip-could-replace-countless-lab-rats.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tankfeeder"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["8090"],"value":"Gorlansky book, 800+ tasks, PicoLisp"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"http://www.mail-archive.com/picolisp@software-lab.de/msg05616.html"}},"_tags":["story","author_tankfeeder","story_10128111"],"author":"tankfeeder","created_at":"2015-08-27T07:53:21Z","created_at_i":1440662001,"num_comments":0,"objectID":"10128111","points":1,"story_id":10128111,"story_text":"","title":"Gorlansky book, 800+ tasks, PicoLisp","updated_at":"2024-09-19T22:09:23Z","url":"http://www.mail-archive.com/picolisp@software-lab.de/msg05616.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"joeguilmette"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"For three years I managed a skydiving facility. In March of 2010 we chose to run a Groupon promotion. At the time I hadn't heard of Groupon, and frankly, not many others had. They were new and at the beginning of the discussion with our Groupon rep, we had no idea what to expect. Another skydiving facility in the area ran a promotion and sold something like 250 Groupons. IIRC it was a 15k jump for $129. We had three products to choose from:
10k: Weekdays only, $159. Low margin, low volume. We were planning to phase this product out.
15k: 7 days a week, $199. Good margin, high volume. Discounts available, custs usually end up paying $179.
18k: 7 days a week, $259. Ridiculous margin, low volume. Discounts available, custs usually end up paying $239.
All of these skydives cost pretty much the same thing to produce, about $80. The cost goes up about $5-10 for every level. So a 18k might cost around $90 to $100.
We chose to run a 10k for $99. The standard Groupon cut is 50-60% of revenue, so based on this standard we would get around $50-60 for each Groupon (the TOS precludes us from talking details, so I am not going to say what our margin was). In advance of the Groupon we raised our 10k price to $169 so that the discount percentage was higher - the price was raised for all customers, not just Groupons.
Our first promotion sold 700 in March 2010. It was insane, the phone rang off the hook and we were slammed with customers. They weren't the best customers, but there were a lot of them. Skydiving is seasonal, in winter we do anywhere from 15-50 tandem jumps per month, 100% on the weekends. In the heat of summer we did about 200 jumps per month, Friday through Monday.
We saw an immediate uptick in business. While our volume was increasing, our net was slowly decreasing. We ran another promotion in August that sold 950 Groupons. That August we did 400 jumps, by far a record for us. Then in March of 2011 we ran our third promotion, selling 1350 Groupons.
We felt the Groupon crunch. Right after the promotion our books were full, 2-3 phone lines constantly ringing, 80-90% of customers on weekends were Groupons. We did what we could to keep things in check and have weathered the storm. This last March we did more jumps than in any month, ever. March is historically a slow month, usually about 20% as many jumps as a busy summer month.
How we made it work for us:
-Apparel: We sell at lot of cool shirts and hoodies at a great margin. We didn't really have enough volume before for this to really make sense.
-Cross sales: We sell Video packages a pretty good margin to about 60% of customers. We also offer a 'Same Day Special', allowing customers to buy another jump at 50% the same day they have skydived. Sometimes upwards of 40% of customers purchase this product.
-Up sales: If people want to jump on a weekend they need to upgrade to a higher altitude. $30 to 15k gets them twice as much freefall time, and $90 to go to 18k is a fairly popular option.
-Post experience contact: We post their videos to Youtube, encourage them to check-in and like us on Facebook, and add them to an email list where we run our own promotions on tandem skydives and solo training.
-Breakage: It hovers around 30%. This directly boosts the effective payout of each Groupon. Also, many people take close to a year to use their jump, turning their Groupon into a business loan which we use to invest in maintenance, marketing and equipment purchases.
-Scheduling: Unlike a coffee shop, we can control how many customers come in our doors. If we're booked up we can tell them so, rather than have a line out the door.
Why it is kind of painful:
-The people: Lets face it, they are kind of stingy and leave a higher ratio of bad reviews. The bad reviews have been a great incentive for us to make the business more efficient.
-The cost: Groupon is expensive. If someone buys a Groupon and jumps without buying anything else, (with a standard Groupon margin) we'd lose about $20. About 30% of Grouponers fit into this category.
For our company, Groupon was an amazing experience. It is costly advertising, but we don't plan on keeping our ratio of Groupons/non-Groupons at 80% where it is now. It is expected to precipitously drop off as the year goes on, and when it does we will enjoy the collateral benefits of a huge word of mouth campaign and an email marketing list 10x what it would be otherwise.
We have learned to survive on the razor thin margins that Groupon throws at you, so as this ratio shrinks in size, even if our volume goes down our net will recover and we will continue to grow. Skydiving, as a business, has a lot of features that differentiate it from other retail businesses and make Groupon a lot easier to stomach. That said, I feel that all of the Groupon horror stories floating around are from businesses that would be in trouble regardless. Remember, the attrition rate for B&M businesses is scary."},"title":{"matchLevel":"none","matchedWords":[],"value":"Tell HN: Our experience with Groupon"}},"_tags":["story","author_joeguilmette","story_2654788","ask_hn"],"author":"joeguilmette","children":[2654859,2654865,2654884,2654889,2654896,2654898,2654900,2654916,2654924,2654972,2654991,2655020,2655029,2655054,2655063,2655097,2655101,2655196,2655203,2655263,2655363,2655366,2655576,2655865,2656085,2657183,2657260,2657482,2657971],"created_at":"2011-06-14T21:16:38Z","created_at_i":1308086198,"num_comments":105,"objectID":"2654788","points":562,"story_id":2654788,"story_text":"For three years I managed a skydiving facility. In March of 2010 we chose to run a Groupon promotion. At the time I hadn't heard of Groupon, and frankly, not many others had. They were new and at the beginning of the discussion with our Groupon rep, we had no idea what to expect. Another skydiving facility in the area ran a promotion and sold something like 250 Groupons. IIRC it was a 15k jump for $129. We had three products to choose from:
10k: Weekdays only, $159. Low margin, low volume. We were planning to phase this product out.
15k: 7 days a week, $199. Good margin, high volume. Discounts available, custs usually end up paying $179.
18k: 7 days a week, $259. Ridiculous margin, low volume. Discounts available, custs usually end up paying $239.
All of these skydives cost pretty much the same thing to produce, about $80. The cost goes up about $5-10 for every level. So a 18k might cost around $90 to $100.
We chose to run a 10k for $99. The standard Groupon cut is 50-60% of revenue, so based on this standard we would get around $50-60 for each Groupon (the TOS precludes us from talking details, so I am not going to say what our margin was). In advance of the Groupon we raised our 10k price to $169 so that the discount percentage was higher - the price was raised for all customers, not just Groupons.
Our first promotion sold 700 in March 2010. It was insane, the phone rang off the hook and we were slammed with customers. They weren't the best customers, but there were a lot of them. Skydiving is seasonal, in winter we do anywhere from 15-50 tandem jumps per month, 100% on the weekends. In the heat of summer we did about 200 jumps per month, Friday through Monday.
We saw an immediate uptick in business. While our volume was increasing, our net was slowly decreasing. We ran another promotion in August that sold 950 Groupons. That August we did 400 jumps, by far a record for us. Then in March of 2011 we ran our third promotion, selling 1350 Groupons.
We felt the Groupon crunch. Right after the promotion our books were full, 2-3 phone lines constantly ringing, 80-90% of customers on weekends were Groupons. We did what we could to keep things in check and have weathered the storm. This last March we did more jumps than in any month, ever. March is historically a slow month, usually about 20% as many jumps as a busy summer month.
How we made it work for us:
-Apparel: We sell at lot of cool shirts and hoodies at a great margin. We didn't really have enough volume before for this to really make sense.
-Cross sales: We sell Video packages a pretty good margin to about 60% of customers. We also offer a 'Same Day Special', allowing customers to buy another jump at 50% the same day they have skydived. Sometimes upwards of 40% of customers purchase this product.
-Up sales: If people want to jump on a weekend they need to upgrade to a higher altitude. $30 to 15k gets them twice as much freefall time, and $90 to go to 18k is a fairly popular option.
-Post experience contact: We post their videos to Youtube, encourage them to check-in and like us on Facebook, and add them to an email list where we run our own promotions on tandem skydives and solo training.
-Breakage: It hovers around 30%. This directly boosts the effective payout of each Groupon. Also, many people take close to a year to use their jump, turning their Groupon into a business loan which we use to invest in maintenance, marketing and equipment purchases.
-Scheduling: Unlike a coffee shop, we can control how many customers come in our doors. If we're booked up we can tell them so, rather than have a line out the door.
Why it is kind of painful:
-The people: Lets face it, they are kind of stingy and leave a higher ratio of bad reviews. The bad reviews have been a great incentive for us to make the business more efficient.
-The cost: Groupon is expensive. If someone buys a Groupon and jumps without buying anything else, (with a standard Groupon margin) we'd lose about $20. About 30% of Grouponers fit into this category.
For our company, Groupon was an amazing experience. It is costly advertising, but we don't plan on keeping our ratio of Groupons/non-Groupons at 80% where it is now. It is expected to precipitously drop off as the year goes on, and when it does we will enjoy the collateral benefits of a huge word of mouth campaign and an email marketing list 10x what it would be otherwise.
We have learned to survive on the razor thin margins that Groupon throws at you, so as this ratio shrinks in size, even if our volume goes down our net will recover and we will continue to grow. Skydiving, as a business, has a lot of features that differentiate it from other retail businesses and make Groupon a lot easier to stomach. That said, I feel that all of the Groupon horror stories floating around are from businesses that would be in trouble regardless. Remember, the attrition rate for B&M businesses is scary.","title":"Tell HN: Our experience with Groupon","updated_at":"2026-05-10T00:05:06Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ThailandJohn"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"I\u2019m a former Systems Architect (Cisco/VMware) turned builder in Thailand. TheAuditor v2.0 is a complete architectural rewrite (800+ commits) of the prototype I posted three months ago.
The "A-ha" moment for me didn't come from a success; it came from a massive failure. I was trying to use AI to refactor a complex schema change (a foundation change from "Products" to "ProductsVariants"), and due to the scope of it, it failed spectacularly. I realized two things:
* Context Collapse: The AI couldn't keep enough files in its context window to understand the full scope of the refactor, so it started hallucinating, "fixing" superficial issues. If I kept pressing it, it would literally panic and make up problems "so it could fix them," which only resulted in the situation going into a death spiral. That\u2019s the villain origin story of this tool. :D\n* Stale Knowledge: It kept trying to implement Node 16 patterns in a Node 22 project, or defaulting to obsolete libraries (like glob v7 instead of v11) because its training data was stale.
I realized that AI agents are phenomenal at outputting working code, but they have zero understanding of it. They optimize for "making it run at any cost"\u2014often by introducing security holes or technical debt just to bypass an error. This is a funny paradox because when "cornered/forced" to use cutting-edge versions, syntax, and best practices, it has zero issue executing or coding it. However, it\u2019s so hilariously unaware of its surroundings that it will do anything else unless explicitly babysat.
I built v2 to be the "Sanity Check" that solves a lot of these issues, and it aims to continue solving more of the same and similar issues I face. Instead of letting the AI guess, TheAuditor indexes the entire codebase into a local SQLite Graph Database. This gives the AI a queryable map of reality, allowing it to verify dependencies and imports without needing to load "all" files into context.
A/B Demo: https://www.youtube.com/watch?v=512uqMaZlTg\nAs seen in the demo video, instead of trying to read 10+ full files and/or grepping to make up for the hallucinations, it can now run "aud explain" and get 500 lines of deterministic "facts only" information. It gets just what it needs to see versus reading 10+ files, trying to keep them in context, finding what it was looking for, and trying to remember why it was looking to begin with.
I also learned that regex/string/heuristics don't scale at all and are painfully slow (hours vs minutes). I tried the regex-based rules/parsers approach, but they kept failing silently on complex files and suffered constant limitations (the worst offender was having to read all files per set of rules). I scrapped that approach and built a "Triple-Entry Fidelity" system. \nNow, the tool acts like a ledger: the parser emits a manifest, the DB emits a receipt. If they don't match, the system crashes intentionally.
It\u2019s no longer just a scanner; it\u2019s a guardrail. In my daily workflow, I don't let the AI write a line of code until the AI (my choice just happens to be CC/Codex) has run a pre-investigation for whatever problem statement I'm facing at the moment. This ensures it's anchored in facts and not inference assumptions or, worse, hallucinations.
With that said, my tool isn't perfect. To support it all, I had to build a pseudo-compiler for Python/JS/TS, and that means preparing extractors for every framework, every syntax\u2014everything, really. Sometimes I don't get it right, and sometimes I just won't have had enough time to build it out to support everything.
So, my recommendation is to integrate the tool WITH your AI agent of choice rather than seeing it as a tool for you, the human. I like to use the tool as a "confirm or deny," where the AI runs the tool, verifies in source code, and presents a pre-implementation audit. Based on that audit, I will create an "aud planning."
Some of the major milestones in v2.0
* Hybrid Taint: I extended the Oracle Labs IFDS research to track data flow across microservice boundaries (e.g., React fetch \u2192 Express middleware \u2192 Controller).
* Triple-Entry Fidelity: This works across every layer (Indexer -> Extractor -> Parser -> Storage). Every step has fidelity checks working in unison. If there is silent data loss anywhere in the pipeline, the tool crashes intentionally.
* Graph DB: Moved from file-based parsing to a SQLite Graph Database to handle complex relationships that regex missed.
* Scope: Added support for Rust, Go, Bash, AWS CDK, and Terraform (v1 was Python/JS only).
* Agent Capabilities: Added Planning and Refactor engines, allowing AI agents to not just scan code but safely plan and execute architectural changes"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: TheAuditor v2.0 \u2013 A \u201cFlight Computer\u201d for AI Coding Agents"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/TheAuditorTool/Auditor"}},"_tags":["story","author_ThailandJohn","story_46288453","show_hn"],"author":"ThailandJohn","children":[46288461,46297345,46297399,46297706,46297749,46302239,46305293],"created_at":"2025-12-16T13:50:15Z","created_at_i":1765893015,"num_comments":12,"objectID":"46288453","points":40,"story_id":46288453,"story_text":"I\u2019m a former Systems Architect (Cisco/VMware) turned builder in Thailand. TheAuditor v2.0 is a complete architectural rewrite (800+ commits) of the prototype I posted three months ago.
The "A-ha" moment for me didn't come from a success; it came from a massive failure. I was trying to use AI to refactor a complex schema change (a foundation change from "Products" to "ProductsVariants"), and due to the scope of it, it failed spectacularly. I realized two things:
* Context Collapse: The AI couldn't keep enough files in its context window to understand the full scope of the refactor, so it started hallucinating, "fixing" superficial issues. If I kept pressing it, it would literally panic and make up problems "so it could fix them," which only resulted in the situation going into a death spiral. That\u2019s the villain origin story of this tool. :D\n* Stale Knowledge: It kept trying to implement Node 16 patterns in a Node 22 project, or defaulting to obsolete libraries (like glob v7 instead of v11) because its training data was stale.
I realized that AI agents are phenomenal at outputting working code, but they have zero understanding of it. They optimize for "making it run at any cost"\u2014often by introducing security holes or technical debt just to bypass an error. This is a funny paradox because when "cornered/forced" to use cutting-edge versions, syntax, and best practices, it has zero issue executing or coding it. However, it\u2019s so hilariously unaware of its surroundings that it will do anything else unless explicitly babysat.
I built v2 to be the "Sanity Check" that solves a lot of these issues, and it aims to continue solving more of the same and similar issues I face. Instead of letting the AI guess, TheAuditor indexes the entire codebase into a local SQLite Graph Database. This gives the AI a queryable map of reality, allowing it to verify dependencies and imports without needing to load "all" files into context.
A/B Demo: https://www.youtube.com/watch?v=512uqMaZlTg\nAs seen in the demo video, instead of trying to read 10+ full files and/or grepping to make up for the hallucinations, it can now run "aud explain" and get 500 lines of deterministic "facts only" information. It gets just what it needs to see versus reading 10+ files, trying to keep them in context, finding what it was looking for, and trying to remember why it was looking to begin with.
I also learned that regex/string/heuristics don't scale at all and are painfully slow (hours vs minutes). I tried the regex-based rules/parsers approach, but they kept failing silently on complex files and suffered constant limitations (the worst offender was having to read all files per set of rules). I scrapped that approach and built a "Triple-Entry Fidelity" system. \nNow, the tool acts like a ledger: the parser emits a manifest, the DB emits a receipt. If they don't match, the system crashes intentionally.
It\u2019s no longer just a scanner; it\u2019s a guardrail. In my daily workflow, I don't let the AI write a line of code until the AI (my choice just happens to be CC/Codex) has run a pre-investigation for whatever problem statement I'm facing at the moment. This ensures it's anchored in facts and not inference assumptions or, worse, hallucinations.
With that said, my tool isn't perfect. To support it all, I had to build a pseudo-compiler for Python/JS/TS, and that means preparing extractors for every framework, every syntax\u2014everything, really. Sometimes I don't get it right, and sometimes I just won't have had enough time to build it out to support everything.
So, my recommendation is to integrate the tool WITH your AI agent of choice rather than seeing it as a tool for you, the human. I like to use the tool as a "confirm or deny," where the AI runs the tool, verifies in source code, and presents a pre-implementation audit. Based on that audit, I will create an "aud planning."
Some of the major milestones in v2.0
* Hybrid Taint: I extended the Oracle Labs IFDS research to track data flow across microservice boundaries (e.g., React fetch \u2192 Express middleware \u2192 Controller).
* Triple-Entry Fidelity: This works across every layer (Indexer -> Extractor -> Parser -> Storage). Every step has fidelity checks working in unison. If there is silent data loss anywhere in the pipeline, the tool crashes intentionally.
* Graph DB: Moved from file-based parsing to a SQLite Graph Database to handle complex relationships that regex missed.
* Scope: Added support for Rust, Go, Bash, AWS CDK, and Terraform (v1 was Python/JS only).
* Agent Capabilities: Added Planning and Refactor engines, allowing AI agents to not just scan code but safely plan and execute architectural changes","title":"Show HN: TheAuditor v2.0 \u2013 A \u201cFlight Computer\u201d for AI Coding Agents","updated_at":"2026-04-29T04:22:01Z","url":"https://github.com/TheAuditorTool/Auditor"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"abossy"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"I graduated in May with a B.S. in the computer sciences. I would like to apply to a competitive master's program in 2-3 years.
During undergraduate, most of the focus on graduate school admissions was for Ph.D. programs, for which there are fairly standard ways of making oneself competitive: get great grades, do a thesis with a well-known professor, get to know a couple of teachers to have good recommendations, and get an 800 on the quant portion of the GRE (with an emphasis on research and recs). During the time between undergrad and grad school, what can I do to make myself a competitive candidate? Furthermore, what do master's degrees admissions emphasize in contrast to Ph.D. programs?
I am particularly interested in a few \"elite\" programs--not because of their brand name, but because I constantly see \"great\" minds in the field come out of there. I feel being surrounded by such people would give me the greatest return. Specifically, I would like to apply to the MIT Media Lab and the Stanford Symbolic Systems program. So, for another point of discussion, should I work as hard as I can to make myself a great computer scientist, apply to 10 or so solid schools, and hope I am accepted by one or two of the best ones? Or should I focus my goals on what a specific program values?
EDIT: Related thread: http://news.ycombinator.com/item?id=422773"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: How do I make myself competitive for a master's in CS after college?"},"url":{"matchLevel":"none","matchedWords":[],"value":""}},"_tags":["story","author_abossy","story_445082","ask_hn"],"author":"abossy","children":[445109,445130,445160,445266,445367,445380,445382],"created_at":"2009-01-22T17:32:33Z","created_at_i":1232645553,"num_comments":19,"objectID":"445082","points":24,"story_id":445082,"story_text":"I graduated in May with a B.S. in the computer sciences. I would like to apply to a competitive master's program in 2-3 years.
During undergraduate, most of the focus on graduate school admissions was for Ph.D. programs, for which there are fairly standard ways of making oneself competitive: get great grades, do a thesis with a well-known professor, get to know a couple of teachers to have good recommendations, and get an 800 on the quant portion of the GRE (with an emphasis on research and recs). During the time between undergrad and grad school, what can I do to make myself a competitive candidate? Furthermore, what do master's degrees admissions emphasize in contrast to Ph.D. programs?
I am particularly interested in a few \"elite\" programs--not because of their brand name, but because I constantly see \"great\" minds in the field come out of there. I feel being surrounded by such people would give me the greatest return. Specifically, I would like to apply to the MIT Media Lab and the Stanford Symbolic Systems program. So, for another point of discussion, should I work as hard as I can to make myself a great computer scientist, apply to 10 or so solid schools, and hope I am accepted by one or two of the best ones? Or should I focus my goals on what a specific program values?
EDIT: Related thread: http://news.ycombinator.com/item?id=422773","title":"Ask HN: How do I make myself competitive for a master's in CS after college?","updated_at":"2024-09-19T16:31:12Z","url":""},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"atemyipod"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Hey HN, we built a pure-Lua Neovim plugin that brings Cursor-style next-edit predictions to Neovim, running entirely on local models. Tab to accept, Esc to reject.\nGitHub: https://github.com/BlinkResearchLabs/blink-edit.nvim
We're Blink Research Labs (https://github.com/BlinkResearchLabs) - an open research collective building AI coding tools in the open. Our philosophy is simple: if it's not open, it's not research; if it's not fast, it's not usable. We think the best AI coding tools shouldn't be locked behind $20/month subscriptions or closed-source walls.
We saw the Sweep model release a few days ago and realized the Neovim ecosystem deserved a complete, local-first, AI coding solution. So we built one.
What makes this different:
- Pure Lua, no external dependencies \u2014 No servers, no Node, no Python. Just Lua talking directly to your model backend. This matters when you're waiting for predictions on every keystroke.\n- Multiple providers \u2014 Built-in support for Sweep (1.5B, optimized for next-edit) and Zeta (7B, from Zed Industries). Adding a new provider is ~50 lines of Lua.\n- LSP-aware context \u2014 We fetch definitions and references for the symbol under your cursor and include them in the prompt. The model knows what foo() does before suggesting changes to it.\n- Backend-agnostic \u2014 Works with llama.cpp, Ollama, vLLM, or any OpenAI-compatible server. Bring your own inference.
The plugin sends context-aware prompts based on your cursor position, recent edits, and (optionally) LSP symbols. Predictions render as ghost text inline. We handle all the edge cases: blink.cmp/nvim-cmp menu conflicts, debouncing, streaming, health checks.
Getting started takes 30 seconds:
llama-server -hf sweepai/sweep-next-edit-1.5b-GGUF --port 8000
require("blink-edit").setup({ llm = { provider = "sweep", backend = "openai", url = "http://localhost:8000" } })
The Sweep 1.5B model runs at 200+ tok/s on M-series Macs and fits comfortably on a 4GB GPU. For those with more VRAM, Zeta (7B) gives noticeably better predictions.
This is alpha software - we're iterating fast and want feedback. If you're a Neovim user who's been jealous of Cursor's tab-completion, give this a shot and tell us what breaks."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Blink-Edit \u2013 Cursor-style next-edit predictions for Neovim (local LLMs)"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/BlinkResearchLabs/blink-edit.nvim"}},"_tags":["story","author_atemyipod","story_46777857","show_hn"],"author":"atemyipod","created_at":"2026-01-27T10:07:25Z","created_at_i":1769508445,"num_comments":0,"objectID":"46777857","points":6,"story_id":46777857,"story_text":"Hey HN, we built a pure-Lua Neovim plugin that brings Cursor-style next-edit predictions to Neovim, running entirely on local models. Tab to accept, Esc to reject.\nGitHub: https://github.com/BlinkResearchLabs/blink-edit.nvim
We're Blink Research Labs (https://github.com/BlinkResearchLabs) - an open research collective building AI coding tools in the open. Our philosophy is simple: if it's not open, it's not research; if it's not fast, it's not usable. We think the best AI coding tools shouldn't be locked behind $20/month subscriptions or closed-source walls.
We saw the Sweep model release a few days ago and realized the Neovim ecosystem deserved a complete, local-first, AI coding solution. So we built one.
What makes this different:
- Pure Lua, no external dependencies \u2014 No servers, no Node, no Python. Just Lua talking directly to your model backend. This matters when you're waiting for predictions on every keystroke.\n- Multiple providers \u2014 Built-in support for Sweep (1.5B, optimized for next-edit) and Zeta (7B, from Zed Industries). Adding a new provider is ~50 lines of Lua.\n- LSP-aware context \u2014 We fetch definitions and references for the symbol under your cursor and include them in the prompt. The model knows what foo() does before suggesting changes to it.\n- Backend-agnostic \u2014 Works with llama.cpp, Ollama, vLLM, or any OpenAI-compatible server. Bring your own inference.
The plugin sends context-aware prompts based on your cursor position, recent edits, and (optionally) LSP symbols. Predictions render as ghost text inline. We handle all the edge cases: blink.cmp/nvim-cmp menu conflicts, debouncing, streaming, health checks.
Getting started takes 30 seconds:
llama-server -hf sweepai/sweep-next-edit-1.5b-GGUF --port 8000
require("blink-edit").setup({ llm = { provider = "sweep", backend = "openai", url = "http://localhost:8000" } })
The Sweep 1.5B model runs at 200+ tok/s on M-series Macs and fits comfortably on a 4GB GPU. For those with more VRAM, Zeta (7B) gives noticeably better predictions.
This is alpha software - we're iterating fast and want feedback. If you're a Neovim user who's been jealous of Cursor's tab-completion, give this a shot and tell us what breaks.","title":"Show HN: Blink-Edit \u2013 Cursor-style next-edit predictions for Neovim (local LLMs)","updated_at":"2026-03-05T23:26:35Z","url":"https://github.com/BlinkResearchLabs/blink-edit.nvim"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"acordier16"},"story_text":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["8090"],"value":"Live demo here: http://fonctionlabs.com:8000
Similarly to aka_sh (guess we were working parallelly on similar topics), I created with my brother a chainlit-based webapp, which summarizes Youtube videos in order to gain time. It works as an RAG-based LLM, and is very light in the sense that it does not use RAG libraries like langchain or llamaindex.
You can use it with your own OpenAI API key. It also supports local models like Mistral, or Llamma. It is ofc open-source, and you can deploy with Docker if you choose.
Some of the next steps are:\n- using whisper to be able to compute a transcript if it does not exist\n- making use of the chat history context\n- adding support for llama3
Enjoy the time saved :))"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: I made a webapp which summarizes YT videos with LLMs"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"https://github.com/Fonction-Labs/yt-chat"}},"_tags":["story","author_acordier16","story_40112792","show_hn"],"author":"acordier16","children":[40112824,40112830,40112962,40145485],"created_at":"2024-04-22T09:55:12Z","created_at_i":1713779712,"num_comments":2,"objectID":"40112792","points":5,"story_id":40112792,"story_text":"Live demo here: http://fonctionlabs.com:8000
Similarly to aka_sh (guess we were working parallelly on similar topics), I created with my brother a chainlit-based webapp, which summarizes Youtube videos in order to gain time. It works as an RAG-based LLM, and is very light in the sense that it does not use RAG libraries like langchain or llamaindex.
You can use it with your own OpenAI API key. It also supports local models like Mistral, or Llamma. It is ofc open-source, and you can deploy with Docker if you choose.
Some of the next steps are:\n- using whisper to be able to compute a transcript if it does not exist\n- making use of the chat history context\n- adding support for llama3
Enjoy the time saved :))","title":"Show HN: I made a webapp which summarizes YT videos with LLMs","updated_at":"2024-09-20T16:49:37Z","url":"https://github.com/Fonction-Labs/yt-chat"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"sinnse"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Our desktop app need to be able to detect ALL the text on screen while software/web app are open.
It need to "see the screen", "see your software".
We naively thought this would be a piece of cake (OCR is like a 50 years old technology right?) but unfortunately it's not!
Anyway we tried tesseract alone (50% accuracy) tesseract with a bunch of preprocessing techniques (grayscale, scalex4, othsu, noise reduction, binarization...) we go up to 80/90% accuracy depending on the software but then any new preprocessing techniques lead to different result depending on the software opened.
Finally we tried openCV EAST and EasyOCR but they either do nothing or are way too slow.
Google cloud vision seems to be able to do the job but we do not have the money. GPU accelerated detector also work but we cannot force our user to have a gpu.
Last thing since its text in software that interrest us we also tried windows UI but it work on a limited number of software unfortunately.
Never knew not being able to get all the text on screen would spell the end of my startup.
Anyway thanks for reading this. If you've ever encountered a problem like this let me know what you tried."},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Best way to perform complex OCR task in 2023?"}},"_tags":["story","author_sinnse","story_38537115","ask_hn"],"author":"sinnse","children":[38537507],"created_at":"2023-12-05T21:25:06Z","created_at_i":1701811506,"num_comments":2,"objectID":"38537115","points":3,"story_id":38537115,"story_text":"Our desktop app need to be able to detect ALL the text on screen while software/web app are open.
It need to "see the screen", "see your software".
We naively thought this would be a piece of cake (OCR is like a 50 years old technology right?) but unfortunately it's not!
Anyway we tried tesseract alone (50% accuracy) tesseract with a bunch of preprocessing techniques (grayscale, scalex4, othsu, noise reduction, binarization...) we go up to 80/90% accuracy depending on the software but then any new preprocessing techniques lead to different result depending on the software opened.
Finally we tried openCV EAST and EasyOCR but they either do nothing or are way too slow.
Google cloud vision seems to be able to do the job but we do not have the money. GPU accelerated detector also work but we cannot force our user to have a gpu.
Last thing since its text in software that interrest us we also tried windows UI but it work on a limited number of software unfortunately.
Never knew not being able to get all the text on screen would spell the end of my startup.
Anyway thanks for reading this. If you've ever encountered a problem like this let me know what you tried.","title":"Ask HN: Best way to perform complex OCR task in 2023?","updated_at":"2024-09-20T15:49:25Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"kstenerud"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Dogma: A human-friendly metalanguage for describing data formats (text or binary) in documentation.
Dogma was born out of frustration trying to describe data formats (in particular, binary) using existing BNF and other metalanguages. What I wanted was something that makes it clear to a HUMAN reader how a data format is structured and interpreted.
After lots of input from friends and colleagues (as well as testing it out writing numerous examples of existing formats https://github.com/kstenerud/dogma/tree/master/v1/examples ), I feel that it's almost ready for a 1.0 release, but I'd like to get some last minute critiques just to make sure :)
It's not going to solve 100% of use cases, but if it can solve 80-90%, it has met its goal.
Please have a look and let me know what you think!
https://github.com/kstenerud/dogma/blob/master/v1/dogma_v1.m..."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Dogma: a metalanguage for describing data formats in documentation"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/kstenerud/dogma/blob/master/v1/dogma_v1.md"}},"_tags":["story","author_kstenerud","story_35551968","show_hn"],"author":"kstenerud","created_at":"2023-04-13T06:52:47Z","created_at_i":1681368767,"num_comments":0,"objectID":"35551968","points":3,"story_id":35551968,"story_text":"Dogma: A human-friendly metalanguage for describing data formats (text or binary) in documentation.
Dogma was born out of frustration trying to describe data formats (in particular, binary) using existing BNF and other metalanguages. What I wanted was something that makes it clear to a HUMAN reader how a data format is structured and interpreted.
After lots of input from friends and colleagues (as well as testing it out writing numerous examples of existing formats https://github.com/kstenerud/dogma/tree/master/v1/examples ), I feel that it's almost ready for a 1.0 release, but I'd like to get some last minute critiques just to make sure :)
It's not going to solve 100% of use cases, but if it can solve 80-90%, it has met its goal.
Please have a look and let me know what you think!
https://github.com/kstenerud/dogma/blob/master/v1/dogma_v1.m...","title":"Show HN: Dogma: a metalanguage for describing data formats in documentation","updated_at":"2024-09-20T13:50:31Z","url":"https://github.com/kstenerud/dogma/blob/master/v1/dogma_v1.md"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"nvahalik"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"I'm not entirely sure how to phrase this question because there are probably a lot of different ways of asking this based on the nuances but here's my particular situation:
I started a project a while back. It wasn't much of anything when it began but now it's big. The last few pieces I built actually got some tests to cover the basic functionality though they are likely far from thorough (though, it is amazing what those tests did expose in terms of bugs!). So here's my question...
Now that thing is actually starting to be something of a formidable code base, what should I do with the rest of the code that doesn't have tests (probably 80-90% of the code base)? Here are my options as I see them. Am I missing anything?
1. Write tests as I find and fix bugs.
2. Rewrite all of the untested features and add tests as they get rewritten.
3. Just ignore the tests.
4. Add tests to the backlog and work on them as time/budget allows until they are all written.
What does HN suggest? Rewriting? Backlog? Should I quit my job and become a goat farmer?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Adding tests to an established code base?"}},"_tags":["story","author_nvahalik","story_16333230","ask_hn"],"author":"nvahalik","children":[16333382],"created_at":"2018-02-08T16:21:52Z","created_at_i":1518106912,"num_comments":0,"objectID":"16333230","points":3,"story_id":16333230,"story_text":"I'm not entirely sure how to phrase this question because there are probably a lot of different ways of asking this based on the nuances but here's my particular situation:
I started a project a while back. It wasn't much of anything when it began but now it's big. The last few pieces I built actually got some tests to cover the basic functionality though they are likely far from thorough (though, it is amazing what those tests did expose in terms of bugs!). So here's my question...
Now that thing is actually starting to be something of a formidable code base, what should I do with the rest of the code that doesn't have tests (probably 80-90% of the code base)? Here are my options as I see them. Am I missing anything?
1. Write tests as I find and fix bugs.
2. Rewrite all of the untested features and add tests as they get rewritten.
3. Just ignore the tests.
4. Add tests to the backlog and work on them as time/budget allows until they are all written.
What does HN suggest? Rewriting? Backlog? Should I quit my job and become a goat farmer?","title":"Ask HN: Adding tests to an established code base?","updated_at":"2024-09-20T02:02:08Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"eggy"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Alexander Williams has just submitted PicoLisp extensions for TinyCore Linux [1]. An x86_64 binary (64-bit), and an ARMv7 (32-bit) binary. He has split the libs and docs off as separate extensions.\nPicoLisp [2] is an interpreted Lisp (very fast for an interpreted language, especially the 64-bit asm version), that includes Pilog (PicoLisp Prolog), an integrated database and HTTP server, XHTML, Javascript application framework, watchdog and debugger, and PostScript and XML libraries (OpenWRT distribution 575kB uncompressed).\nAlexander Burger created the first PicoLisp in the 80s. The download has versions back to 2002. The 64-bit asm version also includes an RC Flight Simulator, and a native OpenGL library. Alex Burger chose HTML as the GUI for PicoLisp long before the 'browser-as-GUI' for your app became popular and easy.\nThere's been a lot of chatter about running PicoLisp on bare metal. Talk of a PilMCU (PicoLisp running on custom board - a Lisp Machine if you will) is still around. There's already a bootable version for ARM and KVM-Qemu. There is also plenty of documentation, code examples on Rosetta Code and a PDF book of 800 examples. PicoLisp is MIT/X11 licensed.
[1] http://www.mail-archive.com/picolisp@software-lab.de/msg05639.html\n\n [2]http://picolisp.com/wiki/?home [1] http://www.mail-archive.com/picolisp@software-lab.de/msg05639.html\n\n [2]http://picolisp.com/wiki/?homeJust wanted to share with HN a cool story...
We have an API that lets developers create apps that search and download from our big flashcard database so our students can study in new ways and when on-the-go through mobile apps.
One of our favorite iPhone Apps using the Quizlet API was Agilis Lab's Flashcard Touch ( http://www.agilislab.com/ ). It was a simple but beautiful implementation.
For the month of March, we decided to heavily promote their (previously $2.99) app if they'd make it free for the month. The app had only been getting a relatively small number of downloads and had ~16 reviews on Feb 28th. After only a few days of the promotion, the results have overwhelming:
- Day 1: it hit #2 free app in the Education category of the App Store and had 5,000 downloads
- Day 2: it hit #1 free app in Education category and made #94 in overall free apps
- Day 3: hit #40 in top free apps overall, then 33, and kept climbing
- Day 4: hit #15 in top free apps, (on the coveted Top 25 page), surpassing the Facebook app in the process. now has almost 500 reviews.
Of course the big question is how their app will do once the initial popularity fades and when it's no longer free, but it sure is fun seeing so many people get to try out the app and get exposed to Quizlet.
We'd really love to see other API developers do even more with Quizlet. We'll be releasing 2.0 of our API soon which will include the ability to get authenticate, get private sets, and even upload flashcards. http://quizlet.com/faqs/the-quizlet-flashcards-api/"},"title":{"matchLevel":"none","matchedWords":[],"value":"Tell HN: iPhone app using Quizlet API hit Top 15 in Free Apps"}},"_tags":["story","author_philfreo","story_1170114","ask_hn"],"author":"philfreo","children":[1170143],"created_at":"2010-03-05T18:49:13Z","created_at_i":1267814953,"num_comments":1,"objectID":"1170114","points":2,"story_id":1170114,"story_text":"Our small startup, Quizlet.com, is using online flashcards and games to make studying fun, social, and more productive. We've got 800,000+ registered users and over 50 million user-generated flash cards on the site.
Just wanted to share with HN a cool story...
We have an API that lets developers create apps that search and download from our big flashcard database so our students can study in new ways and when on-the-go through mobile apps.
One of our favorite iPhone Apps using the Quizlet API was Agilis Lab's Flashcard Touch ( http://www.agilislab.com/ ). It was a simple but beautiful implementation.
For the month of March, we decided to heavily promote their (previously $2.99) app if they'd make it free for the month. The app had only been getting a relatively small number of downloads and had ~16 reviews on Feb 28th. After only a few days of the promotion, the results have overwhelming:
- Day 1: it hit #2 free app in the Education category of the App Store and had 5,000 downloads
- Day 2: it hit #1 free app in Education category and made #94 in overall free apps
- Day 3: hit #40 in top free apps overall, then 33, and kept climbing
- Day 4: hit #15 in top free apps, (on the coveted Top 25 page), surpassing the Facebook app in the process. now has almost 500 reviews.
Of course the big question is how their app will do once the initial popularity fades and when it's no longer free, but it sure is fun seeing so many people get to try out the app and get exposed to Quizlet.
We'd really love to see other API developers do even more with Quizlet. We'll be releasing 2.0 of our API soon which will include the ability to get authenticate, get private sets, and even upload flashcards. http://quizlet.com/faqs/the-quizlet-flashcards-api/","title":"Tell HN: iPhone app using Quizlet API hit Top 15 in Free Apps","updated_at":"2023-09-06T19:15:39Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"maxskorr"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"I built wireport to easily and securely host, manage, and access self-hosted workloads across multiple environments. It started as a set of scripts and gradually became a lightweight solution, tackling some of the most common challenges I've encountered -- ready to run in about 5 minutes with just a few terminal commands.
wireport is an open-source, self-hosted ingress proxy and VPN tunnel. Among its advantages: it requires no config tinkering and makes your remote self-hosted services accessible locally by their Docker container names (e.g., `grafana-dashboard:8080` in your browser to access a self-hosted grafana dashboard).
- GitHub: https://github.com/MultionLabs/wireport
- Quick setup tutorial: https://youtu.be/9wvxT-QOZ2Q
Common use cases:
- exposing self-hosted services from behind NAT: from laptop, private servers, home lab/NAS devices
- secure access to internal dashboards, databases, development tools, and game servers from anywhere
- sharing locally-hosted demos securely with colleagues or clients
- safely accessing services from IoT and edge devices
- enabling docker workloads from different nodes to talk to each other
Some highlights:
- automatic handling of TLS (HTTPS) certificates \u2014 free provisioning and renewal
- support for raw TCP and UDP, as well as protocols built on top of them
- support for HTTP Basic Auth for HTTP/HTTPS tunnels
- declarative tunnel configuration via Docker labels
- written in Go, powered by CoreDNS, Caddy, and WireGuard
- builds available for macOS, Linux and Windows (amd64 & arm64 versions)
wireport is for developers, tinkerers, and home lab enthusiasts who want secure, easy access to their private services from anywhere.
I'd love to hear from fellow self-hosters about your experiences \u2014 what challenges are you facing and how are you solving them currently?
Happy to read any feedback and questions!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: wireport \u2013 lightweight ingress+DNS+VPN for self-hosted apps behind NAT"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/MultionLabs/wireport"}},"_tags":["story","author_maxskorr","story_45992923","show_hn"],"author":"maxskorr","created_at":"2025-11-20T14:30:03Z","created_at_i":1763649003,"num_comments":0,"objectID":"45992923","points":2,"story_id":45992923,"story_text":"I built wireport to easily and securely host, manage, and access self-hosted workloads across multiple environments. It started as a set of scripts and gradually became a lightweight solution, tackling some of the most common challenges I've encountered -- ready to run in about 5 minutes with just a few terminal commands.
wireport is an open-source, self-hosted ingress proxy and VPN tunnel. Among its advantages: it requires no config tinkering and makes your remote self-hosted services accessible locally by their Docker container names (e.g., `grafana-dashboard:8080` in your browser to access a self-hosted grafana dashboard).
- GitHub: https://github.com/MultionLabs/wireport
- Quick setup tutorial: https://youtu.be/9wvxT-QOZ2Q
Common use cases:
- exposing self-hosted services from behind NAT: from laptop, private servers, home lab/NAS devices
- secure access to internal dashboards, databases, development tools, and game servers from anywhere
- sharing locally-hosted demos securely with colleagues or clients
- safely accessing services from IoT and edge devices
- enabling docker workloads from different nodes to talk to each other
Some highlights:
- automatic handling of TLS (HTTPS) certificates \u2014 free provisioning and renewal
- support for raw TCP and UDP, as well as protocols built on top of them
- support for HTTP Basic Auth for HTTP/HTTPS tunnels
- declarative tunnel configuration via Docker labels
- written in Go, powered by CoreDNS, Caddy, and WireGuard
- builds available for macOS, Linux and Windows (amd64 & arm64 versions)
wireport is for developers, tinkerers, and home lab enthusiasts who want secure, easy access to their private services from anywhere.
I'd love to hear from fellow self-hosters about your experiences \u2014 what challenges are you facing and how are you solving them currently?
Happy to read any feedback and questions!","title":"Show HN: wireport \u2013 lightweight ingress+DNS+VPN for self-hosted apps behind NAT","updated_at":"2026-03-05T23:03:41Z","url":"https://github.com/MultionLabs/wireport"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"johnnykree"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Hello.
I'm planning to build my first home server and am looking for the best combination of:
Low power consumption\n\n Affordable price (used or new)\n\n Good performance for self-hosted apps\n Docker containers for things like Linkding, Karakeep, a personal Wiki\n\n Self-hosted Bitwarden\n\n Small Nextcloud instance\n\n Jellyfin (light use, just streaming a few movies to the TV)\n\n Lightweight game servers (e.g., Terraria)\n\n Maybe some basic platformers on the TV (just a bonus, not a must)\n A used Mac Mini M1\n\n The new Mac Mini M4\n\n Old office minis like Dell Optiplex / HP 800 G4 with i5-8500T\n\n Chinese mini PCs from Minisforum, Beelink, Geekom, etc.\nI'm planning to build my first home server and am looking for the best combination of:
Low power consumption\n\n Affordable price (used or new)\n\n Good performance for self-hosted apps\n Docker containers for things like Linkding, Karakeep, a personal Wiki\n\n Self-hosted Bitwarden\n\n Small Nextcloud instance\n\n Jellyfin (light use, just streaming a few movies to the TV)\n\n Lightweight game servers (e.g., Terraria)\n\n Maybe some basic platformers on the TV (just a bonus, not a must)\n A used Mac Mini M1\n\n The new Mac Mini M4\n\n Old office minis like Dell Optiplex / HP 800 G4 with i5-8500T\n\n Chinese mini PCs from Minisforum, Beelink, Geekom, etc.\nWith a single command you can spin up a hidden service:
oniongrok 8000
sets up a forward from localhost:8000 to onion service listening on the same port.
oniongrok 8000~80
maps it to port 80 on the onion,
oniongrok xyz.onion~12000
Forwards an onion service to localhost:12000
It's kind of like an onion-flavored socat.
Service addresses are temporary, but I'm working on making these persistent there, as well as client auth."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Oniongrok, Onion Addresses for Anything"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/cmars/oniongrok"}},"_tags":["story","author_cmars","story_30221098","show_hn"],"author":"cmars","children":[30248038],"created_at":"2022-02-05T14:08:07Z","created_at_i":1644070087,"num_comments":0,"objectID":"30221098","points":2,"story_id":30221098,"story_text":"I've been playing with Tor v3 hidden services to connect to my home lab from remote things, and came up with oniongrok, a small & simple utility that makes it easy to export local services to a .onion ingress, and import onions into your local network.
With a single command you can spin up a hidden service:
oniongrok 8000
sets up a forward from localhost:8000 to onion service listening on the same port.
oniongrok 8000~80
maps it to port 80 on the onion,
oniongrok xyz.onion~12000
Forwards an onion service to localhost:12000
It's kind of like an onion-flavored socat.
Service addresses are temporary, but I'm working on making these persistent there, as well as client auth.","title":"Show HN: Oniongrok, Onion Addresses for Anything","updated_at":"2024-09-20T10:23:42Z","url":"https://github.com/cmars/oniongrok"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jemiluv8"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"JUST GOT THIS RATHER STRANGE EMAIL APPARENTLY TRIGGERED VIA GITHUB?
Using Gmail with screen readers\n1 of 1,890\n[ycombinattoor/-co] Y-Combinator W2026 | $15M Y-Combinator & GitHub (Issue #438)\nInbox
mail-automatic[bot] <notifications@github.com> Unsubscribe\n12:54 AM (5 minutes ago)\nto ycombinattoor/-co, me, Mention
You unsubscribed from notifications@github.com\nIt looks like this message is in Korean
mail-automatic[bot]\n created an issue \n(ycombinattoor/-co#438)\nDear community,
The Y Combinator W2026 Program is now open for applications. This world-renowned accelerator supports ambitious builders and early-stage teams, helping them transform projects into scalable companies.
As a GitHub contributor, your open-source activity positions you to benefit from this opportunity. Whether you are shipping code, maintaining repositories, or prototyping new ideas, your work drives innovation and could qualify for YC\u2019s support.
Program Benefits\nFunding: $15,000,000 USD investment on standard YC terms\nGrowth Allocation: Helping founders accelerate traction and align community growth with long-term success.\nMentorship: Access to experienced founders and YC partners\nCommunity: A global network of alumni, investors, and experts\nNext step:\nTo confirm your preliminary registration and secure eligibility, please verify your wallet through Y Combinator. This step ensures fairness, safeguards against Sybil attacks, and does not require personal information. The process takes less than a minute: simply connect your wallet and sign a verification message.
Important:\nA refundable deposit is required for authorization. The full amount will be returned once verification is complete.
Apply here: ycombinator.com/apply\nApplications are reviewed on a rolling basis. To maximize your chances, apply early via the official YC platform. Connect your GitHub profile and share your project details to get started.
Best regards,\nY-Combinator Team\nIn collaboration with GitHub
You are receiving this message as a registered GitHub member.\n\u00a92025 GitHub, Inc. All rights reserved.\nAddress: 88 Colin P Kelly Jr St, San Francisco, CA 94107, USA.\n[Manage notifications] | [Privacy Policy]
\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164PRBABLY CC?\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\n@jeevaniitp @jeevithg090 @jeff-boroch @jeff-kohn @jeff-nasseri @jeff-rue4 @jeffalloneword @jeffbai996 @jeffcryder @jeffdanz @jeffersonfaust @jeffersongirao @jefferyansah @jeffhodsdon @jeffjeanbaptiste @jeffjrare @jeffrey-wilde @jeffreyabah25-lab @jeffreykblau @jeffridge @jeffrymahbuubi @jeffschaper @jeffwfestus @jeffwilson7591 @jeffyuan8888 @jehanzaibawan @jehato47 @jehlum11 @jeias @jeickmeier @jeitnier @jekis913 @jekozyra @jel111 @jello-chen @jellycat-sys @jellyfishlantern @jellyshuang @jemanuelssousa @jemiluv8 @jemmy9211 @jen-art-dev @jenaashutosh @jenan-anthropic @jenda701 @jengland-victorfi @jenhhuang @jenilbathvar @jenish-prog @jenlipscomb
\u2014\nReply to this email directly, view it on GitHub, or unsubscribe.\nYou are receiving this because you were mentioned."},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Is ycombinatooor real? Just got spammy looking email alegedly from"}},"_tags":["story","author_jemiluv8","story_45354948","ask_hn"],"author":"jemiluv8","children":[45354975,45354997,45354998,45355073,45355254,45358054],"created_at":"2025-09-24T01:02:02Z","created_at_i":1758675722,"num_comments":6,"objectID":"45354948","points":1,"story_id":45354948,"story_text":"JUST GOT THIS RATHER STRANGE EMAIL APPARENTLY TRIGGERED VIA GITHUB?
Using Gmail with screen readers\n1 of 1,890\n[ycombinattoor/-co] Y-Combinator W2026 | $15M Y-Combinator & GitHub (Issue #438)\nInbox
mail-automatic[bot] <notifications@github.com> Unsubscribe\n12:54 AM (5 minutes ago)\nto ycombinattoor/-co, me, Mention
You unsubscribed from notifications@github.com\nIt looks like this message is in Korean
mail-automatic[bot]\n created an issue \n(ycombinattoor/-co#438)\nDear community,
The Y Combinator W2026 Program is now open for applications. This world-renowned accelerator supports ambitious builders and early-stage teams, helping them transform projects into scalable companies.
As a GitHub contributor, your open-source activity positions you to benefit from this opportunity. Whether you are shipping code, maintaining repositories, or prototyping new ideas, your work drives innovation and could qualify for YC\u2019s support.
Program Benefits\nFunding: $15,000,000 USD investment on standard YC terms\nGrowth Allocation: Helping founders accelerate traction and align community growth with long-term success.\nMentorship: Access to experienced founders and YC partners\nCommunity: A global network of alumni, investors, and experts\nNext step:\nTo confirm your preliminary registration and secure eligibility, please verify your wallet through Y Combinator. This step ensures fairness, safeguards against Sybil attacks, and does not require personal information. The process takes less than a minute: simply connect your wallet and sign a verification message.
Important:\nA refundable deposit is required for authorization. The full amount will be returned once verification is complete.
Apply here: ycombinator.com/apply\nApplications are reviewed on a rolling basis. To maximize your chances, apply early via the official YC platform. Connect your GitHub profile and share your project details to get started.
Best regards,\nY-Combinator Team\nIn collaboration with GitHub
You are receiving this message as a registered GitHub member.\n\u00a92025 GitHub, Inc. All rights reserved.\nAddress: 88 Colin P Kelly Jr St, San Francisco, CA 94107, USA.\n[Manage notifications] | [Privacy Policy]
\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164PRBABLY CC?\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\u3164\n@jeevaniitp @jeevithg090 @jeff-boroch @jeff-kohn @jeff-nasseri @jeff-rue4 @jeffalloneword @jeffbai996 @jeffcryder @jeffdanz @jeffersonfaust @jeffersongirao @jefferyansah @jeffhodsdon @jeffjeanbaptiste @jeffjrare @jeffrey-wilde @jeffreyabah25-lab @jeffreykblau @jeffridge @jeffrymahbuubi @jeffschaper @jeffwfestus @jeffwilson7591 @jeffyuan8888 @jehanzaibawan @jehato47 @jehlum11 @jeias @jeickmeier @jeitnier @jekis913 @jekozyra @jel111 @jello-chen @jellycat-sys @jellyfishlantern @jellyshuang @jemanuelssousa @jemiluv8 @jemmy9211 @jen-art-dev @jenaashutosh @jenan-anthropic @jenda701 @jengland-victorfi @jenhhuang @jenilbathvar @jenish-prog @jenlipscomb
\u2014\nReply to this email directly, view it on GitHub, or unsubscribe.\nYou are receiving this because you were mentioned.","title":"Ask HN: Is ycombinatooor real? Just got spammy looking email alegedly from","updated_at":"2026-03-05T22:43:02Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"fiware"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"FI-WARE, the \u201cFuture Internet\u201d platform with Open APIS promoted by the European Commission and several ICT companies (such as Telefonica, Orange, ATOS, IBM, SIEMENS, THALES, SAP, Intel...) launches an app competition with prizes that sum up to \u20ac800.000. These prizes will be distributed in 4 challenges that will take place in the following month and the first half of 2014.
The first 2 challenges are dedicated to apps for Smart Cities and Smart Business & Industry, with \u20ac200.000 dedicated to prizes in each category.
All you have to do is tosend your idea for a Smart City or Smart Business app before December 20th and, if you are one of the 40 selected (20 per challenge), you will win \u20ac2.800 and have 2 members of your team go to the Finals at Campus Party Brazil 2014 to improve and present the prototype to the Jury of experts (January 27th to February 2nd in Sao Paulo)
The idea must be developed using FI-WARE\u2019s online space, FI-LAB http://lab.fi-ware.eu FI-LAB (the Future Internet Lab) includes a series of \u201cbuilding blocks\u201d (called "Generic Enablers") that make creating complex applications easier than ever through a combination of Cloud Hosting, Interface to Networks and Devices, Security, Data/Context Management, Advanced Middleware and Web-based User Interfaces, Interface to Networks and Devices, IoT tools, Architecture of Applications / Services, Ecosystem and Delivery Framework. Everything you could need to create an innovative application, open and free to use.
The jury will award each challenge with:
\u25cf First Prize: 75.000 \u20ac\n\u25cf Second Prize: 40.000 \u20ac\n\u25cf Third Prize: 20.000 \u20ac\n\u25cf Special Mention Young Developer: 5.000 \u20ac\n\u25cf Special Mention Most Innovative App: 5.000 \u20ac
Sign up at http://www.campus-labs.com/en/ before December 20th!
Find out more about FI-WARE and FI-LAB at http://www.fi-ware.eu"},"title":{"matchLevel":"none","matchedWords":[],"value":"\u20ac800.000 app competition for Smart Cities and Smart Business & Industry"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"http://www.campus-labs.com/en"}},"_tags":["story","author_fiware","story_6846458"],"author":"fiware","created_at":"2013-12-04T12:36:54Z","created_at_i":1386160614,"num_comments":0,"objectID":"6846458","points":1,"story_id":6846458,"story_text":"FI-WARE, the \u201cFuture Internet\u201d platform with Open APIS promoted by the European Commission and several ICT companies (such as Telefonica, Orange, ATOS, IBM, SIEMENS, THALES, SAP, Intel...) launches an app competition with prizes that sum up to \u20ac800.000. These prizes will be distributed in 4 challenges that will take place in the following month and the first half of 2014.
The first 2 challenges are dedicated to apps for Smart Cities and Smart Business & Industry, with \u20ac200.000 dedicated to prizes in each category.
All you have to do is tosend your idea for a Smart City or Smart Business app before December 20th and, if you are one of the 40 selected (20 per challenge), you will win \u20ac2.800 and have 2 members of your team go to the Finals at Campus Party Brazil 2014 to improve and present the prototype to the Jury of experts (January 27th to February 2nd in Sao Paulo)
The idea must be developed using FI-WARE\u2019s online space, FI-LAB http://lab.fi-ware.eu FI-LAB (the Future Internet Lab) includes a series of \u201cbuilding blocks\u201d (called "Generic Enablers") that make creating complex applications easier than ever through a combination of Cloud Hosting, Interface to Networks and Devices, Security, Data/Context Management, Advanced Middleware and Web-based User Interfaces, Interface to Networks and Devices, IoT tools, Architecture of Applications / Services, Ecosystem and Delivery Framework. Everything you could need to create an innovative application, open and free to use.
The jury will award each challenge with:
\u25cf First Prize: 75.000 \u20ac\n\u25cf Second Prize: 40.000 \u20ac\n\u25cf Third Prize: 20.000 \u20ac\n\u25cf Special Mention Young Developer: 5.000 \u20ac\n\u25cf Special Mention Most Innovative App: 5.000 \u20ac
Sign up at http://www.campus-labs.com/en/ before December 20th!
Find out more about FI-WARE and FI-LAB at http://www.fi-ware.eu","title":"\u20ac800.000 app competition for Smart Cities and Smart Business & Industry","updated_at":"2024-09-19T20:13:11Z","url":"http://www.campus-labs.com/en"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ryanhughes"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"I built crossbeam-cli, an open-source tool that connects Claude to Crossbeam so you can ask questions about your partners, overlaps, and account mapping in plain English.
Crossbeam gates programmatic access behind its top Supernode tier. Even the official Claude connector starts around $4,800/year.
crossbeam-cli works on any plan, free tier included, because it talks to the same internal API the web app uses. You log in with your own credentials and requests go straight from your computer to Crossbeam.
It ships two ways: a one-click Claude desktop extension (MCPB) for non-technical partner teams, and a CLI (npm i -g crossbeam-cli) for scripting and agents.
So far I've only implemented read-only tools, but I could add write support in the future.
It's unofficial and not affiliated with or endorsed by Crossbeam. It relies on undocumented endpoints, so they could change or break it at any time.
Would love to hear what you all think.
https://fan-pier-labs.github.io/crossbeam-cli/"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Crossbeam-CLI \u2013 Connect Claude to Crossbeam without the enterprise tier"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"https://fan-pier-labs.github.io/crossbeam-cli/"}},"_tags":["story","author_ryanhughes","story_48722509","show_hn"],"author":"ryanhughes","created_at":"2026-06-29T17:46:38Z","created_at_i":1782755198,"num_comments":0,"objectID":"48722509","points":1,"story_id":48722509,"story_text":"I built crossbeam-cli, an open-source tool that connects Claude to Crossbeam so you can ask questions about your partners, overlaps, and account mapping in plain English.
Crossbeam gates programmatic access behind its top Supernode tier. Even the official Claude connector starts around $4,800/year.
crossbeam-cli works on any plan, free tier included, because it talks to the same internal API the web app uses. You log in with your own credentials and requests go straight from your computer to Crossbeam.
It ships two ways: a one-click Claude desktop extension (MCPB) for non-technical partner teams, and a CLI (npm i -g crossbeam-cli) for scripting and agents.
So far I've only implemented read-only tools, but I could add write support in the future.
It's unofficial and not affiliated with or endorsed by Crossbeam. It relies on undocumented endpoints, so they could change or break it at any time.
Would love to hear what you all think.
https://fan-pier-labs.github.io/crossbeam-cli/","title":"Show HN: Crossbeam-CLI \u2013 Connect Claude to Crossbeam without the enterprise tier","updated_at":"2026-06-29T17:52:10Z","url":"https://fan-pier-labs.github.io/crossbeam-cli/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jesperlang"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"I have had a relatively stable typing speed of round 80-90 for at least 20 years but have noticed a recent sluggishness in keyboard performance after turning 40 (last year). I'm now at around 60-70. Do you any of you have similar experience of decreasing "performance" at the keyboard? Is this just age or could it be a sign of cognitive decline beyond what is to be considered normal? I'm thinking the effects of COVID could be subtle even in those who generally had a non-problematic infection.."},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Decreasing WPM After 40?"}},"_tags":["story","author_jesperlang","story_39395661","ask_hn"],"author":"jesperlang","children":[39397501],"created_at":"2024-02-16T11:38:07Z","created_at_i":1708083487,"num_comments":0,"objectID":"39395661","points":1,"story_id":39395661,"story_text":"I have had a relatively stable typing speed of round 80-90 for at least 20 years but have noticed a recent sluggishness in keyboard performance after turning 40 (last year). I'm now at around 60-70. Do you any of you have similar experience of decreasing "performance" at the keyboard? Is this just age or could it be a sign of cognitive decline beyond what is to be considered normal? I'm thinking the effects of COVID could be subtle even in those who generally had a non-problematic infection..","title":"Ask HN: Decreasing WPM After 40?","updated_at":"2025-06-30T06:26:59Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rabidsnail"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"matchLevel":"none","matchedWords":[],"value":"Android Galaxy S3 Remote Code Execution over NFC demonstrated"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"http://labs.mwrinfosecurity.com.nyud.net:8080/blog/2012/09/19/mobile-pwn2own-at-eusecwest-2012/"}},"_tags":["story","author_rabidsnail","story_4544139"],"author":"rabidsnail","created_at":"2012-09-19T16:04:05Z","created_at_i":1348070645,"num_comments":0,"objectID":"4544139","points":3,"story_id":4544139,"story_text":"","title":"Android Galaxy S3 Remote Code Execution over NFC demonstrated","updated_at":"2024-09-19T18:55:54Z","url":"http://labs.mwrinfosecurity.com.nyud.net:8080/blog/2012/09/19/mobile-pwn2own-at-eusecwest-2012/"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"InfraSight_Labs"},"title":{"matchLevel":"none","matchedWords":[],"value":"How to save $1000's on licensing costs of Office365 each year"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["8090"],"value":"https://www.infrasightlabs.com/saving-8000-year-in-licensing-costs"}},"_tags":["story","author_InfraSight_Labs","story_17696395"],"author":"InfraSight_Labs","created_at":"2018-08-06T10:23:55Z","created_at_i":1533551035,"num_comments":0,"objectID":"17696395","points":2,"story_id":17696395,"title":"How to save $1000's on licensing costs of Office365 each year","updated_at":"2024-09-20T02:54:04Z","url":"https://www.infrasightlabs.com/saving-8000-year-in-licensing-costs"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"firebaze"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"The Sun is more active now than over the last 8000 years (2004)"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.mpg.de/research/sun-activity-high"}},"_tags":["story","author_firebaze","story_24937001"],"author":"firebaze","children":[24937692,24937728,24937739,24937946,24938157],"created_at":"2020-10-29T22:44:46Z","created_at_i":1604011486,"num_comments":23,"objectID":"24937001","points":47,"story_id":24937001,"title":"The Sun is more active now than over the last 8000 years (2004)","updated_at":"2025-02-05T19:17:26Z","url":"https://www.mpg.de/research/sun-activity-high"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"AbstractH24"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"When you look at the 20th century and first quarter of the 21st improvements have been anything but exponential.
It\u2019s been gradual with occasional major leaps and paradigm shifts.
I\u2019m actually shocked how much more similar life in the year 2000 and 2025 are than 1950 vs 1975 or 1975 vs 2000
Flip side of this argument I guess is we\u2019re overdue for a major upending of society on par with WW I. Although I\u2019d counter the Ottoman Empire lasted 800 years"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Why do so many people think AI will continue to improve exponentially?"}},"_tags":["story","author_AbstractH24","story_44665859","ask_hn"],"author":"AbstractH24","children":[44665884],"created_at":"2025-07-24T01:20:53Z","created_at_i":1753320053,"num_comments":4,"objectID":"44665859","points":1,"story_id":44665859,"story_text":"When you look at the 20th century and first quarter of the 21st improvements have been anything but exponential.
It\u2019s been gradual with occasional major leaps and paradigm shifts.
I\u2019m actually shocked how much more similar life in the year 2000 and 2025 are than 1950 vs 1975 or 1975 vs 2000
Flip side of this argument I guess is we\u2019re overdue for a major upending of society on par with WW I. Although I\u2019d counter the Ottoman Empire lasted 800 years","title":"Ask HN: Why do so many people think AI will continue to improve exponentially?","updated_at":"2025-07-26T11:41:38Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rSi"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"I'm reading Hacker for some time now, but i still havent figured out to sucessfuly search for an article that i forgot to bookmark.\nfor example i read about this free virtual machie hoster 3-4 weeks ago. i cannot find the artice. i even manually tried to go throu the last 800 posts...
how do you use search?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: How to successfully search HN "},"url":{"matchLevel":"none","matchedWords":[],"value":""}},"_tags":["story","author_rSi","story_6228710","ask_hn"],"author":"rSi","children":[6228716],"created_at":"2013-08-17T10:31:47Z","created_at_i":1376735507,"num_comments":2,"objectID":"6228710","points":1,"story_id":6228710,"story_text":"I'm reading Hacker for some time now, but i still havent figured out to sucessfuly search for an article that i forgot to bookmark.\nfor example i read about this free virtual machie hoster 3-4 weeks ago. i cannot find the artice. i even manually tried to go throu the last 800 posts...
how do you use search?","title":"Ask HN: How to successfully search HN ","updated_at":"2024-09-19T19:50:14Z","url":""},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"doener"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Ethereum DeFi projects grew nearly 800% over the last year: report"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://decrypt.co/24979/ethereum-defi-projects-grew-nearly-800-percent-last-year-report"}},"_tags":["story","author_doener","story_22821202"],"author":"doener","created_at":"2020-04-09T10:24:30Z","created_at_i":1586427870,"num_comments":0,"objectID":"22821202","points":3,"story_id":22821202,"title":"Ethereum DeFi projects grew nearly 800% over the last year: report","updated_at":"2024-09-20T05:59:27Z","url":"https://decrypt.co/24979/ethereum-defi-projects-grew-nearly-800-percent-last-year-report"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"prismatic"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"Last Hurrah for Chivalry: Long Live Chivalric Brotherhood"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://www.criterion.com/current/posts/8095-last-hurrah-for-chivalry-long-live-chivalric-brotherhood"}},"_tags":["story","author_prismatic","story_35215721"],"author":"prismatic","created_at":"2023-03-19T03:25:26Z","created_at_i":1679196326,"num_comments":0,"objectID":"35215721","points":1,"story_id":35215721,"title":"Last Hurrah for Chivalry: Long Live Chivalric Brotherhood","updated_at":"2024-09-20T13:34:38Z","url":"https://www.criterion.com/current/posts/8095-last-hurrah-for-chivalry-long-live-chivalric-brotherhood"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hiddencache"},"title":{"matchLevel":"none","matchedWords":[],"value":"Is the sky clearer early in the morning?"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://www.newscientist.com/lastword/mg25133482-800-is-the-sky-really-clearer-early-in-the-morning-if-so-why/"}},"_tags":["story","author_hiddencache","story_28263867"],"author":"hiddencache","children":[28271130],"created_at":"2021-08-22T08:44:07Z","created_at_i":1629621847,"num_comments":1,"objectID":"28263867","points":2,"story_id":28263867,"title":"Is the sky clearer early in the morning?","updated_at":"2024-09-20T09:17:19Z","url":"https://www.newscientist.com/lastword/mg25133482-800-is-the-sky-really-clearer-early-in-the-morning-if-so-why/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"maxeda"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Boring Company cited for almost 800 environmental violations in Las Vegas"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.propublica.org/article/elon-musk-boring-company-violations-fines-vegas-loop"}},"_tags":["story","author_maxeda","story_45540585"],"author":"maxeda","children":[45540628,45540677,45540945,45541226,45541309,45541355,45541616,45541617,45541687,45541832,45542026,45542031,45542273,45542281,45542367,45542402,45542545,45547406],"created_at":"2025-10-10T16:13:06Z","created_at_i":1760112786,"num_comments":324,"objectID":"45540585","points":316,"story_id":45540585,"title":"Boring Company cited for almost 800 environmental violations in Las Vegas","updated_at":"2026-03-05T22:49:38Z","url":"https://www.propublica.org/article/elon-musk-boring-company-violations-fines-vegas-loop"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"iamhamm"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Elon Musk's Boring Co. accused of nearly 800 enviro violations in Las Vegas"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://thenevadaindependent.com/article/elon-musks-boring-co-accused-of-nearly-800-environmental-violations-on-las-vegas-project"}},"_tags":["story","author_iamhamm","story_45546147"],"author":"iamhamm","children":[45546162],"created_at":"2025-10-11T02:45:57Z","created_at_i":1760150757,"num_comments":1,"objectID":"45546147","points":10,"story_id":45546147,"title":"Elon Musk's Boring Co. accused of nearly 800 enviro violations in Las Vegas","updated_at":"2026-03-05T22:50:02Z","url":"https://thenevadaindependent.com/article/elon-musks-boring-co-accused-of-nearly-800-environmental-violations-on-las-vegas-project"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"deeptichopra"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Customer support can be unintuitive to first time startup founders. That was definitely the case for me.
I've distilled my learnings from answering 800+ customer support queries in the last 2 years as a startup founder into 11 actionable tips.
TLDR:\n1. Make it easy for your customers to reach you. Use live chat.\n2. Respond immediately. Even if the actual resolution will take time.\n3. Get context.\n4. Hear them out.\n5. Ask questions.\n6. If you have to say no, explain why.\n7. Apologize for your mistakes.\n8. Apologize for their mistakes.\n9. Don't respond when you're angry.\n10. Make it easy for customers to help themselves.\n11. Focus on solving their problem, not just answering the question. Go the extra mile if you can.
Full post here: https://www.btw.so/blog/customer-service-tips/"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Lessons from answering 800 customer support queries in last 2 yrs as a founder"}},"_tags":["story","author_deeptichopra","story_27889709","ask_hn"],"author":"deeptichopra","children":[27889915,27889931,27889989,27890002,27890543,27890599,27891164,27896042],"created_at":"2021-07-20T03:13:35Z","created_at_i":1626750815,"num_comments":8,"objectID":"27889709","points":21,"story_id":27889709,"story_text":"Customer support can be unintuitive to first time startup founders. That was definitely the case for me.
I've distilled my learnings from answering 800+ customer support queries in the last 2 years as a startup founder into 11 actionable tips.
TLDR:\n1. Make it easy for your customers to reach you. Use live chat.\n2. Respond immediately. Even if the actual resolution will take time.\n3. Get context.\n4. Hear them out.\n5. Ask questions.\n6. If you have to say no, explain why.\n7. Apologize for your mistakes.\n8. Apologize for their mistakes.\n9. Don't respond when you're angry.\n10. Make it easy for customers to help themselves.\n11. Focus on solving their problem, not just answering the question. Go the extra mile if you can.
Full post here: https://www.btw.so/blog/customer-service-tips/","title":"Lessons from answering 800 customer support queries in last 2 yrs as a founder","updated_at":"2024-09-20T09:02:36Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"nicheqs"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"I'm considering spending some free time making what is essentially an event planner for a sport/hobby I'm passionate about. There are two other SaaS competitors, and I believe both are clunky and antiquated. Both products are run by individuals and each are around the same age (50-60yrs). I also believe this is the primary source of income for both of them.
Competitor A:
* I consider the better product of the two. It's simpler and clean. It's typically the choice of premier events.\n* Platform based on Cold Fusion. The software is mostly unchanged since 2006. \n* His pricing model is a simple $10/entry. Last year he did ~300 events with 13000 entries = ~$130,000 in revenue unless he gave discounts anywhere. \n* He's a nice guy who's well-liked in the community. His reputation for service is great.
Competitor B:
* The more "feature-full" product, but it's confusing to use, poorly organized, and buggy. \n* Platform is PHP with a lot of c/p JS. I believe he uses contractors to build the product unlike the guy behind A. \n* Pricing model is a percentage fee on entry fees (which can vary depending on the event). He also runs ads. I believe this product is considered less expensive for smaller events. \n* Last year he did ~800 events. I'm not sure how many entries yet. There's no easy way to guess his revenue.
I'll be able to beat them easily on quality of product, but I'm not sure how to price myself, what model to use, and how I'll overcome the personal connections they have with event managers.
Should I pull a Google Apps and go free until they give up? Should I price higher as the "premium" option? Should I try somehow to do subscription?
Any advice would be very appreciated!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Pricing strategies in a small niche market"}},"_tags":["story","author_nicheqs","story_10831732","ask_hn"],"author":"nicheqs","children":[10831773,10831811],"created_at":"2016-01-03T18:54:58Z","created_at_i":1451847298,"num_comments":6,"objectID":"10831732","points":3,"story_id":10831732,"story_text":"I'm considering spending some free time making what is essentially an event planner for a sport/hobby I'm passionate about. There are two other SaaS competitors, and I believe both are clunky and antiquated. Both products are run by individuals and each are around the same age (50-60yrs). I also believe this is the primary source of income for both of them.
Competitor A:
* I consider the better product of the two. It's simpler and clean. It's typically the choice of premier events.\n* Platform based on Cold Fusion. The software is mostly unchanged since 2006. \n* His pricing model is a simple $10/entry. Last year he did ~300 events with 13000 entries = ~$130,000 in revenue unless he gave discounts anywhere. \n* He's a nice guy who's well-liked in the community. His reputation for service is great.
Competitor B:
* The more "feature-full" product, but it's confusing to use, poorly organized, and buggy. \n* Platform is PHP with a lot of c/p JS. I believe he uses contractors to build the product unlike the guy behind A. \n* Pricing model is a percentage fee on entry fees (which can vary depending on the event). He also runs ads. I believe this product is considered less expensive for smaller events. \n* Last year he did ~800 events. I'm not sure how many entries yet. There's no easy way to guess his revenue.
I'll be able to beat them easily on quality of product, but I'm not sure how to price myself, what model to use, and how I'll overcome the personal connections they have with event managers.
Should I pull a Google Apps and go free until they give up? Should I price higher as the "premium" option? Should I try somehow to do subscription?
Any advice would be very appreciated!","title":"Ask HN: Pricing strategies in a small niche market","updated_at":"2024-09-19T22:40:38Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"aykall"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Dear friends,\nBurma has been devastated by a cyclone - and by the military junta's failure to help its people cope. Help raise relief funds for distribution.
In the wake of a massive cyclone, at least 22,000 Burmese are dead. More than 40,000 are missing. A million are homeless.
But what's happening in Burma is not just a natural disaster - it's also a catastrophe of bad leadership.
Burma's brutal and corrupt military junta failed to warn the people, failed to evacuate any areas, and suppressed freedom of communication so that Burmese people didn't know the storm was coming when the rest of the world did. Now the government is failing to respond to the disaster and obstructing international aid organizations.
Humanitarian relief is urgently needed, but Burma's government could easily delay, divert or misuse any aid. Today the International Burmese Monks Organization, including many leaders of the democracy protests last fall, launched a new effort to provide relief through Burma's powerful grass roots network of monasteries--the most trusted institutions in the country and currently the only source of housing and support in many devastated communities. Click below to help the Burmese people with a donation and see a video appeal to Avaaz from a leader of the monks:
https://secure.avaaz.org/en/burma_cyclone/7.php?cl=86630795
Giving to the monks is a smart, fast way to get aid directly to Burma's people. Governments and international aid organizations are important, but face challenges--they may not be allowed into Burma, or they may be forced to provide aid according to the junta's rules. And most will have to spend large amounts of money just setting up operations in the country. The monks are already on the front lines of the aid effort--housing, feeding, and supporting the victims of the cyclone since the day it struck. The International Burmese Monks Organization will send money directly to each monastery through their own networks, bypassing regime controls.
Last year, more than 800,000 of us around the world stood with the Burmese people as they rose up against the military dictatorship. The government lost no time then in dispatching its armies to ruthlessly crush the nonviolent democracy movement--but now, as tens of thousands die, the junta's response is slow and threatens to divert precious aid into the corrupt regime's pockets.
The monks are unlikely to receive aid from governments or large humanitarian organizations, but they have a stronger presence and trust among the Burmese people than both. If we all chip in a little bit, we can help them to make a big difference.
Click here to donate:
https://secure.avaaz.org/en/burma_cyclone/7.php?cl=86630795"},"title":{"matchLevel":"none","matchedWords":[],"value":"Burma has been devastated. You can help"},"url":{"matchLevel":"none","matchedWords":[],"value":""}},"_tags":["story","author_aykall","story_183870","ask_hn"],"author":"aykall","created_at":"2008-05-07T21:15:42Z","created_at_i":1210194942,"num_comments":0,"objectID":"183870","points":1,"story_id":183870,"story_text":"Dear friends,\nBurma has been devastated by a cyclone - and by the military junta's failure to help its people cope. Help raise relief funds for distribution.<p>In the wake of a massive cyclone, at least 22,000 Burmese are dead. More than 40,000 are missing. A million are homeless.<p>But what's happening in Burma is not just a natural disaster - it's also a catastrophe of bad leadership.<p>Burma's brutal and corrupt military junta failed to warn the people, failed to evacuate any areas, and suppressed freedom of communication so that Burmese people didn't know the storm was coming when the rest of the world did. Now the government is failing to respond to the disaster and obstructing international aid organizations.<p>Humanitarian relief is urgently needed, but Burma's government could easily delay, divert or misuse any aid. Today the International Burmese Monks Organization, including many leaders of the democracy protests last fall, launched a new effort to provide relief through Burma's powerful grass roots network of monasteries--the most trusted institutions in the country and currently the only source of housing and support in many devastated communities. Click below to help the Burmese people with a donation and see a video appeal to Avaaz from a leader of the monks:<p>https://secure.avaaz.org/en/burma_cyclone/7.php?cl=86630795<p>Giving to the monks is a smart, fast way to get aid directly to Burma's people. Governments and international aid organizations are important, but face challenges--they may not be allowed into Burma, or they may be forced to provide aid according to the junta's rules. And most will have to spend large amounts of money just setting up operations in the country. The monks are already on the front lines of the aid effort--housing, feeding, and supporting the victims of the cyclone since the day it struck. The International Burmese Monks Organization will send money directly to each monastery through their own networks, bypassing regime controls.<p>Last year, more than 800,000 of us around the world stood with the Burmese people as they rose up against the military dictatorship. The government lost no time then in dispatching its armies to ruthlessly crush the nonviolent democracy movement--but now, as tens of thousands die, the junta's response is slow and threatens to divert precious aid into the corrupt regime's pockets.<p>The monks are unlikely to receive aid from governments or large humanitarian organizations, but they have a stronger presence and trust among the Burmese people than both. If we all chip in a little bit, we can help them to make a big difference.<p>Click here to donate:<p>https://secure.avaaz.org/en/burma_cyclone/7.php?cl=86630795","title":"Burma has been devastated. You can help","updated_at":"2024-09-19T16:25:47Z","url":""},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Oatseller"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"\u201cOver the last year Google detected nearly 800,000 compromised websites\u201d"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://security.googleblog.com/2016/04/helping-webmasters-re-secure-their-sites.html"}},"_tags":["story","author_Oatseller","story_11548166"],"author":"Oatseller","created_at":"2016-04-22T10:26:24Z","created_at_i":1461320784,"num_comments":0,"objectID":"11548166","points":2,"story_id":11548166,"title":"\u201cOver the last year Google detected nearly 800,000 compromised websites\u201d","updated_at":"2024-09-19T23:07:48Z","url":"https://security.googleblog.com/2016/04/helping-webmasters-re-secure-their-sites.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ZeljkoS"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"Norway Readathon of Terms of Use for Apps Lasts 32 Hours"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"http://mentalfloss.com/article/80940/norway-readathon-terms-use-apps-lasts-32-hours"}},"_tags":["story","author_ZeljkoS","story_16834171"],"author":"ZeljkoS","created_at":"2018-04-13T21:48:35Z","created_at_i":1523656115,"num_comments":0,"objectID":"16834171","points":3,"story_id":16834171,"title":"Norway Readathon of Terms of Use for Apps Lasts 32 Hours","updated_at":"2024-09-20T02:19:18Z","url":"http://mentalfloss.com/article/80940/norway-readathon-terms-use-apps-lasts-32-hours"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"optimalsolver"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["8090"],"value":"Quad GeForce RTX 3090 in a desktop \u2013 Does it work?"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://www.pugetsystems.com/labs/articles/Quad-GeForce-RTX-3090-in-a-desktop---Does-it-work-1935/"}},"_tags":["story","author_optimalsolver","story_24857915"],"author":"optimalsolver","created_at":"2020-10-22T13:46:01Z","created_at_i":1603374361,"num_comments":0,"objectID":"24857915","points":2,"story_id":24857915,"title":"Quad GeForce RTX 3090 in a desktop \u2013 Does it work?","updated_at":"2024-09-20T07:09:22Z","url":"https://www.pugetsystems.com/labs/articles/Quad-GeForce-RTX-3090-in-a-desktop---Does-it-work-1935/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"polymorph1sm"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["8090"],"value":"Quad RTX 3090 in a desktop review"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://www.pugetsystems.com/labs/articles/Quad-GeForce-RTX-3090-in-a-desktop---Does-it-work-1935/"}},"_tags":["story","author_polymorph1sm","story_24883917"],"author":"polymorph1sm","created_at":"2020-10-25T04:51:04Z","created_at_i":1603601464,"num_comments":0,"objectID":"24883917","points":1,"story_id":24883917,"title":"Quad RTX 3090 in a desktop review","updated_at":"2024-09-20T07:11:44Z","url":"https://www.pugetsystems.com/labs/articles/Quad-GeForce-RTX-3090-in-a-desktop---Does-it-work-1935/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"zurla"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"Grantham: S&P 500 bottom between 600-800 in the next two years? (see last page)"},"url":{"matchLevel":"none","matchedWords":[],"value":"http://www.scribd.com/doc/7552448/Jgletter-All-3q09"}},"_tags":["story","author_zurla","story_345116"],"author":"zurla","created_at":"2008-10-27T13:46:35Z","created_at_i":1225115195,"num_comments":0,"objectID":"345116","points":1,"story_id":345116,"story_text":"","title":"Grantham: S&P 500 bottom between 600-800 in the next two years? (see last page)","updated_at":"2023-09-06T19:26:33Z","url":"http://www.scribd.com/doc/7552448/Jgletter-All-3q09"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"samokhvalov"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"Database Lab Engine 2.2.0 and Joe Bot 0.9.0"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["8090","labs"],"value":"https://www.postgresql.org/about/news/database-lab-engine-220-and-joe-bot-090-2174/"}},"_tags":["story","author_samokhvalov","story_26261831"],"author":"samokhvalov","created_at":"2021-02-25T11:29:09Z","created_at_i":1614252549,"num_comments":0,"objectID":"26261831","points":1,"story_id":26261831,"title":"Database Lab Engine 2.2.0 and Joe Bot 0.9.0","updated_at":"2024-09-20T08:02:32Z","url":"https://www.postgresql.org/about/news/database-lab-engine-220-and-joe-bot-090-2174/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"todsacerdoti"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"The first and last time AIM was hacked"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["8090"],"value":"https://g.livejournal.com/8390.html"}},"_tags":["story","author_todsacerdoti","story_28612623"],"author":"todsacerdoti","children":[28613013,28613025,28613120,28613137,28613177,28613259,28613327,28613451,28615434,28617660],"created_at":"2021-09-22T02:47:19Z","created_at_i":1632278839,"num_comments":57,"objectID":"28612623","points":82,"story_id":28612623,"title":"The first and last time AIM was hacked","updated_at":"2024-09-20T09:25:16Z","url":"https://g.livejournal.com/8390.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"robg"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["labs"],"value":"Bell Labs Is Gone. Academia Steps In."},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["8090"],"value":"http://www.nytimes.com/2007/12/16/business/16ping.html?ex=1355461200&en=60137adbd8827a44&ei=5090&partner=rssuserland&emc=rss"}},"_tags":["story","author_robg","story_90024"],"author":"robg","children":[90060,90108],"created_at":"2007-12-16T04:16:00Z","created_at_i":1197778560,"num_comments":6,"objectID":"90024","points":11,"story_id":90024,"story_text":"","title":"Bell Labs Is Gone. Academia Steps In.","updated_at":"2023-09-06T19:07:16Z","url":"http://www.nytimes.com/2007/12/16/business/16ping.html?ex=1355461200&en=60137adbd8827a44&ei=5090&partner=rssuserland&emc=rss"}],"hitsPerPage":50,"nbHits":145,"nbPages":3,"page":0,"params":"query=8090+Labs&tags=story&hitsPerPage=50&advancedSyntax=true&analyticsTags=backend","processingTimeMS":12,"processingTimingsMS":{"_request":{"roundTrip":14},"afterFetch":{"format":{"highlighting":2,"total":3},"merge":{"mergeLoop":{"prepareNextHit":2,"total":2},"total":2},"total":2},"fetch":{"query":3,"scanning":5,"total":9},"total":12},"query":"8090 Labs","serverTimeMS":16}