id
int64 2
42.1M
| by
large_stringlengths 2
15
⌀ | time
timestamp[us] | title
large_stringlengths 0
198
⌀ | text
large_stringlengths 0
27.4k
⌀ | url
large_stringlengths 0
6.6k
⌀ | score
int64 -1
6.02k
⌀ | descendants
int64 -1
7.29k
⌀ | kids
large list | deleted
large list | dead
bool 1
class | scraping_error
large_stringclasses 25
values | scraped_title
large_stringlengths 1
59.3k
⌀ | scraped_published_at
large_stringlengths 4
66
⌀ | scraped_byline
large_stringlengths 1
757
⌀ | scraped_body
large_stringlengths 1
50k
⌀ | scraped_at
timestamp[us] | scraped_language
large_stringclasses 58
values | split
large_stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
42,024,661 | edward | 2024-11-02T07:31:00 | SmolLM2 | null | https://simonwillison.net/2024/Nov/2/smollm2/ | 110 | 47 | [
42025116,
42025050,
42025151,
42025092,
42025051,
42025670,
42025414,
42025025,
42025062,
42025180,
42028703,
42025464
] | null | null | null | null | null | null | null | null | null | train |
42,024,670 | neochau | 2024-11-02T07:32:44 | Show HN: I made a tutorial of how to use free edge TTS API with deno.js [video] | null | https://www.youtube.com/watch?v=uATgw1KVj_4 | 5 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,024,680 | _Microft | 2024-11-02T07:34:28 | Practical Radio Circuits (2003) [pdf] | null | https://www.fracassi.net/iw2ntf/manuali/practical-radio-circuits-raymond-haigh.pdf | 6 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,024,692 | cannibalXxx | 2024-11-02T07:37:28 | Apple reaches deal to acquire Pixelmator | null | https://9to5mac.com/2024/11/01/apple-reaches-deal-to-acquire-pixelmator/ | 2 | 0 | null | null | null | no_error | Apple reaches deal to acquire Pixelmator - 9to5Mac | 2024-11-01T15:31:01+00:00 | Chance Miller |
Apple is acquiring the popular photo editing company Pixelmator. The Pixelmator Team announced the news in a blog post on Friday and said there will be “no material changes” to its apps “at this time.” The deal is pending regulatory approval.
Pixelmator has signed an agreement to be acquired by Apple, subject to regulatory approval. There will be no material changes to the Pixelmator Pro, Pixelmator for iOS, and Photomator apps at this time. Stay tuned for exciting updates to come.
Pixelmator, a 17-year-old company, says that the Apple acquisition will let its tools have an even bigger impact on creative people around the world:
We’ve been inspired by Apple since day one, crafting our products with the same razor-sharp focus on design, ease of use, and performance. And looking back, it’s crazy what a small group of dedicated people have been able to achieve over the years from all the way in Vilnius, Lithuania. Now, we’ll have the ability to reach an even wider audience and make an even bigger impact on the lives of creative people around the world.
Pixelmator has become a staple of the Apple ecosystem over the years with its incredibly popular and powerful Pixelmator Pro and Photomator apps. The apps have also added a number of powerful AI and ML features over the years, including background removal tools and Super Resolution.
For those unfamiliar, Pixelmator Pro is described as a “professional image editing [tool] that anyone can use.” Photomator, meanwhile, is the “ultimate photo editor for Mac, iPhone, iPad, and Vision Pro.”
Apple acquires companies on a regular basis, but it’s rare that acquisitions are ever publicly announced like this. One of the more recent examples of a move like this is when Apple acquired Dark Sky in 2020. Apple also acquired Workflow in 2017 ahead of the launch of Shortcuts.
Add 9to5Mac to your Google News feed.
FTC: We use income earning auto affiliate links. More. | 2024-11-08T13:54:45 | en | train |
42,024,703 | alwillis | 2024-11-02T07:39:35 | Apple Expands Globalstar Relationship to Support Growing Satellite Services | null | https://www.macrumors.com/2024/11/01/apple-globalstar-satellite-agreement/ | 5 | 1 | [
42024725,
42025673
] | null | null | null | null | null | null | null | null | null | train |
42,024,727 | jart | 2024-11-02T07:45:26 | Weird Lexical Syntax | null | https://justine.lol/lex/ | 429 | 226 | [
42026536,
42026377,
42031338,
42026212,
42026050,
42026136,
42026421,
42026160,
42028425,
42026417,
42026177,
42026154,
42027870,
42026106,
42026051,
42026459,
42026607,
42026422,
42026555,
42027030,
42026132,
42025949,
42039966,
42030197,
42026782,
42029176,
42027557,
42025850,
42033616,
42026554,
42029276,
42031544,
42026211,
42028642,
42026078,
42025979
] | null | null | null | null | null | null | null | null | null | train |
42,024,740 | mahin | 2024-11-02T07:48:09 | Some Philosophical Problems from the Standpoint of AI (1969) [pdf] | null | http://www-formal.stanford.edu/jmc/mcchay69.pdf | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,024,748 | noduerme | 2024-11-02T07:49:33 | Ask HN: Panicking about new AWS MFA reqs | I've run dozens of servers/services/DBs on AWS as the root user for years. I will never, ever set up MFA there or anywhere else. I <i>want only email verification</i> through my private email server. No hardware keys to lose, and no cell phone to go missing or have to reload my credentials on (my phones disappear a lot). Suddenly the AWS root login tells me MFA will be <i>required</i> within 6 days through either a hardware key or an authenticator app.<p>There seems to be no way to get in touch with Amazon about it, so I'm asking here. Anyone else in the same situation?<p>BTW, Don't tell me just accept it. If you do, you don't get what I'm talking about. I can't get a phone number of who the f** I'd have to talk to there to permanently exempt my account from this req. I've been moving stuff there for 10 years and I'm managing $100k /yr through them at least. Moving verification to an authenticator app or key is MUCH less secure for my living situation. | null | 1 | 9 | [
42024811,
42027514,
42024833
] | null | null | null | null | null | null | null | null | null | train |
42,024,757 | cannibalXxx | 2024-11-02T07:52:12 | How to definitively learn regular expressions in programming languages | To learn regular expressions once and for all, we'll understand what they are, when to use them and how to master them with a few study strategies.<p>What are regular expressions?
Regular expressions (or “regex”) are strings of characters that define search patterns. They are widely used to search, manipulate and validate textual data, either in a line of code or in a data set. They are supported in several programming languages, such as Python, JavaScript, PHP, C++, and many others.<p>In essence, a regular expression is like a search language that allows you to define specific patterns for:<p>- Search for words in a text.<p>- Identify numbers in sequences.<p>- Filter strings based on complex criteria.<p>Use Cases for Regular Expressions
1. Form Validation: Regex is common in validating emails, passwords and other form entries, ensuring that a piece of data follows the right format.<p>2. Text Search and Replace: Replace strings or parts of strings in a document or file based on patterns. Example: changing all telephone numbers to a single format.<p>3. Data Extraction: Extract complex data from text, such as URLs, IP addresses, hashtags or mentions on social networks.
4. Log Analysis: Useful for analyzing server logs, where regexes can help filter out specific messages or errors.<p>5. Web Scraping: When capturing information from web pages, regular expressions help to search for structured data in HTML.<p>6. Structured Data Parsing: Manipulate or restructure JSON, XML, CSV etc. data based on specific standards.<p>How to Learn Regular Expressions the Definitive Way
1- Understand the Basic Structure:<p>Start by learning the main operators and characters:
- . (dot) - any character.<p>- *, +, ? - repetition quantifiers.<p>- [abc], [^abc] - character classes.<p>- \d, \w, \s - shortcut metacharacters.<p>- ^, $ - start and end of line anchors.<p>2- Practice with Interactive Tools:<p>- Tools like [Regex101](https://regex101.com/) or [Regexr](https://regexr.com/) help you test expressions in real time, see detailed explanations and understand how they work in different languages.<p>3- Learn to Read and Write Complex Regex in Parts:<p>- Divide more complex expressions into blocks. Practice reading and understanding each block before combining them.<p>4- Solve practical problems:<p>- Apply regexes to your own projects, such as filtering emails in a contact list or formatting dates. Practicing in real-life scenarios helps to solidify your learning.<p>5- Revise Regularly and Gradually Increase Complexity:<p>- Revise regularly and always look for new challenges, evolving from basic expressions to more advanced regexes. | null | 2 | 5 | [
42024848,
42025681
] | null | null | null | null | null | null | null | null | null | train |
42,024,794 | kolinko | 2024-11-02T08:00:01 | Does Godwin's law (rule of Nazi analogies) apply in observable reality? | null | https://cyber.harvard.edu/story/2021-12/does-godwins-law-rule-nazi-analogies-apply-observable-reality-empirical-study | 4 | 2 | [
42025030,
42024813
] | null | null | null | null | null | null | null | null | null | train |
42,024,802 | cranberryturkey | 2024-11-02T08:01:58 | Ask HN: Does anyone have any successes with Jobsolv.com? | I'm debating if its worth the money. | null | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,024,810 | benithemaker | 2024-11-02T08:04:12 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,024,821 | tobr | 2024-11-02T08:06:59 | 3D Workers Island | null | https://www.yes-thread.com/3dwiscr/ | 4 | 1 | [
42024931
] | null | null | null | null | null | null | null | null | null | train |
42,024,822 | zamiir | 2024-11-02T08:06:59 | Guutaale | null | https://blog.cloudflare.com/post-mortem-todays-attack-apparent-google-app/ | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,024,823 | muthuh | 2024-11-02T08:07:05 | Python #1 on GitHub and speedup via Codon / Cython | null | https://hwisnu.bearblog.dev/python-1-on-github-speedup-via-codon-cython/ | 2 | 0 | [
42024824
] | null | null | null | null | null | null | null | null | null | train |
42,024,825 | hjuutilainen | 2024-11-02T08:07:19 | Broadcom Brings Back VSphere Enterprise Plus and VSphere Standard | null | https://news.broadcom.com/cloud/bringing-more-value-and-options-to-the-vmware-cloud-foundation-portfolio | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,024,843 | null | 2024-11-02T08:12:18 | null | null | null | null | null | null | [
"true"
] | null | null | null | null | null | null | null | null | train |
42,024,849 | skeememe | 2024-11-02T08:14:18 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,024,864 | martinrue | 2024-11-02T08:17:46 | What If You Hadn't? | null | https://martinrue.com/what-if-you-hadn%27t | 2 | 4 | [
42024910
] | null | null | null | null | null | null | null | null | null | train |
42,024,872 | shscs911 | 2024-11-02T08:19:46 | Four Eras of JavaScript Frameworks (2022) | null | https://www.pzuraq.com/blog/four-eras-of-javascript-frameworks | 3 | 1 | [
42024973
] | null | null | null | null | null | null | null | null | null | train |
42,024,874 | epilys | 2024-11-02T08:19:56 | The Pixel Aesthetic (2017) | null | https://www.alexhw.com/pixelAesthetic.html | 2 | 0 | [
42025669
] | null | null | null | null | null | null | null | null | null | train |
42,024,882 | AlexeyKrivov | 2024-11-02T08:21:08 | Making SaaS Friendly Figma UI Kit | null | https://www.glowui.com/ | 3 | 1 | [
42024883,
42025671
] | null | null | null | null | null | null | null | null | null | train |
42,024,886 | droidrat | 2024-11-02T08:21:50 | null | null | null | 1 | null | [
42024887
] | null | true | null | null | null | null | null | null | null | train |
42,024,899 | xbmcuser | 2024-11-02T08:24:49 | A startup is acquiring enormous diamonds–and using blockchain tech to shake up | null | https://fortune.com/2024/10/31/diamond-jewelry-blockchain-hb-antwerp-de-beers-botswana/ | 2 | 1 | [
42024923
] | null | null | null | null | null | null | null | null | null | train |
42,024,919 | mahin | 2024-11-02T08:29:14 | Australia's weird little time zone | null | https://www.howderfamily.com/blog/australias-weird-little-time-zone/ | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,024,925 | Pavlinbg | 2024-11-02T08:31:38 | null | null | null | 1 | null | [
42024926
] | null | true | null | null | null | null | null | null | null | train |
42,024,930 | shinryuu | 2024-11-02T08:33:22 | Unidentified Floating Object: Edo Images of UFOs | null | https://publicdomainreview.org/collection/utsuro-bune/ | 3 | 3 | [
42024971
] | null | null | null | null | null | null | null | null | null | train |
42,024,947 | nkko | 2024-11-02T08:37:57 | Foundational Principles for AI Development Environments | null | https://pulent.com/posts/ai-dev-envs/ | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,024,960 | networked | 2024-11-02T08:41:14 | w3c/png: Consider using Zstandard and/or LZ4 instead of Deflate (2021) | null | https://github.com/w3c/png/issues/39 | 3 | 0 | [
42025662,
42025665
] | null | null | null | null | null | null | null | null | null | train |
42,024,962 | bizlinkers | 2024-11-02T08:42:13 | null | null | null | 1 | null | [
42024963
] | null | true | null | null | null | null | null | null | null | train |
42,024,976 | withakay | 2024-11-02T08:46:40 | DevOps is a Culture not a Function (isn't it?) | null | https://badsector.dev/posts/devops-is-a-culture-not-a-function/ | 3 | 1 | [
42025169,
42024977,
42025660
] | null | null | null | null | null | null | null | null | null | train |
42,024,989 | noteable | 2024-11-02T08:49:38 | With the help of OpenAI, the website was developed and launched in just one day | null | https://www.stripetop.com/ | 1 | 0 | [
42024990
] | null | null | null | null | null | null | null | null | null | train |
42,024,996 | lifeisstillgood | 2024-11-02T08:51:32 | Most Small Businesses do not make a profit | null | https://www.youtube.com/watch?v=s9ta9bXB5lg | 1 | 1 | [
42025427,
42025401
] | null | null | null | null | null | null | null | null | null | train |
42,025,014 | markgavalda | 2024-11-02T08:55:49 | Apple Researchers Show Critical Flaw in AI | null | https://www.latimes.com/business/story/2024-11-01/column-these-apple-researchers-just-proved-that-ai-bots-cant-think-and-possibly-never-will | 12 | 8 | [
42025089,
42025794,
42025384,
42025656,
42025399
] | null | null | no_error | Hiltzik: Apple ponders what AI is good for - Los Angeles Times | 2024-11-01T10:00:09.895Z | Michael Hiltzik | See if you can solve this arithmetic problem:Oliver picks 44 kiwis on Friday. Then he picks 58 kiwis on Saturday. On Sunday, he picks double the number of kiwis he did on Friday, but five of them were a bit smaller than average. How many kiwis does Oliver have?If you answered “190,” congratulations: You did as well as the average grade school kid by getting it right. (Friday’s 44 plus Saturday’s 58 plus Sunday’s 44 multiplied by 2, or 88, equals 190.) You also did better than more than 20 state-of-the-art artificial intelligence models tested by an AI research team at Apple. The AI bots, they found, consistently got it wrong. The fact that Apple did this has gotten a lot of attention, but nobody should be surprised at the results. — AI critic Gary Marcus The Apple team found “catastrophic performance drops” by those models when they tried to parse simple mathematical problems written in essay form. In this example, the systems tasked with the question often didn’t understand that the size of the kiwis have nothing to do with the number of kiwis Oliver has. Some, consequently, subtracted the five undersized kiwis from the total and answered “185.” Human schoolchildren, the researchers posited, are much better at detecting the difference between relevant information and inconsequential curveballs. Get the latest from Michael Hiltzik Commentary on economics and more from a Pulitzer Prize winner. You may occasionally receive promotional content from the Los Angeles Times. The Apple findings were published in October in a technical paper that has attracted widespread attention in AI labs and the lay press, not only because the results are well-documented, but also because the researchers work for the nation’s leading high-tech consumer company — and one that has just rolled out a suite of purported AI features for iPhone users. “The fact that Apple did this has gotten a lot of attention, but nobody should be surprised at the results,” says Gary Marcus, a critic of how AI systems have been marketed as reliably, well, “intelligent.”Indeed, Apple’s conclusion matches earlier studies that have found that large language models, or LLMs, don’t actually “think” so much as match language patterns in materials they’ve been fed as part of their “training.” When it comes to abstract reasoning — “a key aspect of human intelligence,” in the words of Melanie Mitchell, an expert in cognition and intelligence at the Santa Fe Institute — the models fall short.“Even very young children are adept at learning abstract rules from just a few examples,” Mitchell and colleagues wrote last year after subjecting GPT bots to a series of analogy puzzles. Their conclusion was that “a large gap in basic abstract reasoning still remains between humans and state-of-the-art AI systems.”That’s important because LLMs such as GPT undergird the AI products that have captured the public’s attention. But the LLMs tested by the Apple team were consistently misled by the language patterns they were trained on. The Apple researchers set out to answer the question, “Do these models truly understand mathematical concepts?” as one of the lead authors, Mehrdad Farajtabar, put it in a thread on X. Their answer is no. They also pondered whether the shortcomings they identified can be easily fixed, and their answer is also no: “Can scaling data, models, or compute fundamentally solve this?” Farajtabar asked in his thread. “We don’t think so!”The Apple research, along with other findings about the limitations of AI bots’ cogitative limitations, is a much-needed corrective to the sales pitches coming from companies hawking their AI models and systems, including OpenAI and Google’s DeepMind lab. The promoters generally depict their products as dependable and their output as trustworthy. In fact, their output is consistently suspect, posing a clear danger when they’re used in contexts where the need for rigorous accuracy is absolute, say in healthcare applications.That’s not always the case. “There are some problems which you can make a bunch of money on without having a perfect solution,” Marcus told me. Recommendation engines powered by AI — those that steer buyers on Amazon to products they might also like, for example. If those systems get a recommendation wrong, it’s no big deal; a customer might spend a few dollars on a book he or she didn’t like. “But a calculator that’s right only 85% of the time is garbage,” Marcus says. “You wouldn’t use it.”The potential for damagingly inaccurate outputs is heightened by AI bots’ natural language capabilities, with which they offer even absurdly inaccurate answers with convincingly cocksure elan. Often they double down on their errors when challenged. These errors are typically described by AI researchers as “hallucinations.” The term may make the mistakes seem almost innocuous, but in some applications, even a minuscule error rate can have severe ramifications. That’s what academic researchers concluded in a recently published analysis of Whisper, an AI-powered speech-to-text tool developed by OpenAI, which can be used to transcribe medical discussions or jailhouse conversations monitored by correction officials. The researchers found that about 1.4% of Whisper-transcribed audio segments in their sample contained hallucinations, including the addition to transcribed conversation of wholly fabricated statements including portrayals of “physical violence or death ... [or] sexual innuendo,” and demographic stereotyping. That may sound like a minor flaw, but the researchers observed that the errors could be incorporated in official records such as transcriptions of court testimony or prison phone calls — which could lead to official decisions based on “phrases or claims that a defendant never said.”Updates to Whisper in late 2023 improved its performance, the researchers said, but the updated Whisper “still regularly and reproducibly hallucinated.”That hasn’t deterred AI promoters from unwarranted boasting about their products. In an Oct. 29 tweet, Elon Musk invited followers to submit “x-ray, PET, MRI or other medical images to Grok [the AI application for his X social media platform] for analysis.” Grok, he wrote, “is already quite accurate and will become extremely good.” It should go without saying that, even if Musk is telling the truth (not an absolutely certain conclusion), any system used by healthcare providers to analyze medical images needs to be a lot better than “extremely good,” however one might define that standard.That brings us to the Apple study. It’s proper to note that the researchers aren’t critics of AI as such but believers that its limitations need to be understood. Farajtabar was formerly a senior research scientist at DeepMind, where another author interned under him; other co-authors hold advanced degrees and professional experience in computer science and machine learning.The team plied their subject AI models with questions drawn from a popular collection of more than 8,000 grade school arithmetic problems testing schoolchildren’s understanding of addition, subtraction, multiplication and division. When the problems incorporated clauses that might seem relevant but weren’t, the models’ performance plummeted. That was true of all the models, including versions of the GPT bots developed by OpenAI, Meta’s Llama, Microsoft’s Phi-3, Google’s Gemma and several models developed by the French lab Mistral AI. Some did better than others, but all showed a decline in performance as the problems became more complex. One problem involved a basket of school supplies including erasers, notebooks and writing paper. That requires a solver to multiply the number of each item by its price and add them together to determine how much the entire basket costs. When the bots were also told that “due to inflation, prices were 10% cheaper last year,” the bots reduced the cost by 10%. That produces a wrong answer, since the question asked what the basket would cost now, not last year.Why did this happen? The answer is that LLMs are developed, or trained, by feeding them huge quantities of written material scraped from published works or the internet — not by trying to teach them mathematical principles. LLMs function by gleaning patterns in the data and trying to match a pattern to the question at hand. But they become “overfitted to their training data,” Farajtabar explained via X. “They memorized what is out there on the web and do pattern matching and answer according to the examples they have seen. It’s still a [weak] type of reasoning but according to other definitions it’s not a genuine reasoning capability.” (the brackets are his.)That’s likely to impose boundaries on what AI can be used for. In mission-critical applications, humans will almost always have to be “in the loop,” as AI developers say—vetting answers for obvious or dangerous inaccuracies or providing guidance to keep the bots from misinterpreting their data, misstating what they know, or filling gaps in their knowledge with fabrications.To some extent, that’s comforting, for it means that AI systems can’t accomplish much without having human partners at hand. But it also means that we humans need to be aware of the tendency of AI promoters to overstate their products’ capabilities and conceal their limitations. The issue is not so much what AI can do, but how users can be gulled into thinking what it can do.“These systems are always going to make mistakes because hallucinations are inherent,” Marcus says. “The ways in which they approach reasoning are an approximation and not the real thing. And none of this is going away until we have some new technology.” More to Read | 2024-11-08T06:15:54 | en | train |
42,025,034 | eftdrone | 2024-11-02T09:02:44 | null | null | null | 1 | null | [
42025035
] | null | true | null | null | null | null | null | null | null | train |
42,025,037 | marklit | 2024-11-02T09:03:28 | 131M American Buildings | null | https://tech.marksblogg.com/ornl-fema-buildings.html | 61 | 10 | [
42071465,
42071126,
42070640,
42025080
] | null | null | null | null | null | null | null | null | null | train |
42,025,055 | fagnerbrack | 2024-11-02T09:07:25 | Exploring Generative AI | null | https://martinfowler.com/articles/exploring-gen-ai.html | 2 | 0 | null | null | null | body_too_long | null | null | null | null | 2024-11-08T13:42:28 | null | train |
42,025,063 | austinallegro | 2024-11-02T09:10:12 | World's Largest captive crocodile dies aged 110 | null | https://news.sky.com/story/worlds-largest-captive-crocodile-dies-aged-110-13246456 | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,081 | walterbell | 2024-11-02T09:16:01 | Autoliv NV2 Night Vision Automotive Camera Teardown (2018) | null | https://debugmo.de/2018/12/autoliv-nv2-teardown/ | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,104 | ChinaTalk | 2024-11-02T09:19:53 | null | null | null | 1 | null | [
42025105
] | null | true | null | null | null | null | null | null | null | train |
42,025,123 | peutetre | 2024-11-02T09:24:57 | Rivian's chief software officer says in-car buttons are 'an anomaly' | null | https://techcrunch.com/2024/10/30/rivians-chief-software-officer-says-in-car-buttons-are-an-anomaly/ | 21 | 86 | [
42025186,
42025286,
42025171,
42025373,
42025336,
42025239,
42025179,
42025251,
42025589,
42025332,
42025241,
42027721,
42025272,
42025391,
42025500,
42025575,
42027864,
42025192,
42029944,
42025379,
42025390,
42025627,
42038833,
42025810,
42025511,
42034966,
42025475,
42025339,
42025238,
42025437,
42025439,
42028622,
42025168,
42025185,
42025298,
42025639,
42025450
] | null | null | null | null | null | null | null | null | null | train |
42,025,138 | nerdyadventurer | 2024-11-02T09:28:25 | Scientists Discover 'Deep Brain' Genes Linked to Parkinson's and ADHD | null | https://www.sciencealert.com/scientists-discover-deep-brain-genes-linked-to-parkinsons-and-adhd | 2 | 1 | [
42025230
] | null | null | null | null | null | null | null | null | null | train |
42,025,142 | alianinfo | 2024-11-02T09:29:24 | Newsletters for DevOps, SRE and Platform Engineers | null | https://lablabs.io/20-of-the-best-newsletters-for-devops-sre-and-platform-engineers/ | 2 | 0 | [
42025638
] | null | null | null | null | null | null | null | null | null | train |
42,025,152 | mrtedbear | 2024-11-02T09:32:29 | the US island ruled by alien snakes and spiders | null | https://www.bbc.com/future/article/20241030-the-island-ruled-by-alien-snakes-and-spiders | 3 | 0 | null | null | null | no_error | The US island ruled by alien snakes and spiders | 2024-10-31T10:00:00.000Z | Zaria Gorvett | Alamy(Credit: Alamy)Guam has 40 times more spiders than neighbouring islands – and a population of invasive snakes so voracious, they have emptied the forests of every bird.Five years ago, Haldre Rogers attended a get-together on the island of Guam – an emerald-green smudge in the western Pacific Ocean, around 2,492km (1,548 miles) from the Philippines. But soon the party was interrupted by an uninvited guest.It was late evening, and outside there was a hog roasting – the remains of dinner. The fire was going down, though still warm. Everyone briefly walked off to chat. When they came back, there was a brown form curled around the pig – something shiny and scaly, with vertical slit-eyes and a wide, smiling mouth. The creature was ripping off chunks of the pig's flesh and swallowing them whole – slowly gulping them into its pale, distended body."It wasn't [exactly] a 400-pound (181kg) pig, but it was a pig for a big party," says Rogers, an associate professor in the Department of Fish and Wildlife Conservation at Virginia Tech in the US, who has been studying Guam's ecology for the last 22 years.AlamyBrown tree snakes will readily eat animals that are already dead – an unusual habit for snakes, which mostly only eat prey they have killed themselves (Credit: Alamy)The visitor without table manners was a brown tree snake – an alien invader which is thought to have been inadvertently introduced to Guam in the 1940s, perhaps after sneaking onto a cargo ship. Before this, an abundance of native birds had enjoyed an idyllic existence in the island's otherworldly limestone forests. But within just four decades of the snake's debut, these voracious predators had begun emptying the jungle of every single one. Out of 12 species, 10 are now extinct on the island, while the remaining two are clinging on in inaccessible caves and urban areas.Now that the avian community has been virtually wiped out, Guam's population of some two million snakes – no one really knows how many there are – will hoover up whatever they can find, including rats, shrews, lizards, or, on this occasion, human leftovers."I mean, they'll eat anything," says Henry Pollock, executive director of the Southern Plains Land Trust, a non-profit in Colorado, who has previously studied the ecology of Guam. "They'll eat each other."With a plague of ravenous snakes, and forests that are entirely devoid of the friendly whistles, chirps and caws that once dominated their soundscape, Guam has become notorious as one of the planet's most spectacular ecological debacles. But the consequences of the island's snake infestation extend far beyond its eerily silent, bird-less forests. Here, an evolutionary experiment is unfolding. And one notable beneficiary has eight legs, plenty of eyes – and the good luck to find itself on an island where sharp, hungry beaks are little more than a distant memory.A web of websRogers is not frightened of spiders – and that's just as well.On most of the Mariana Islands, there are relatively few spiders in the rainy season, with a large spike as the climate dries out. But not on Guam. The island's limestone forests are an arachno-nightmare all year round – a near-continuous tangle of silvery threads that stretches for miles, where every step you take reveals another web and its hairy host.There are giant, yellow-bellied banana spiders, with their golden webs in the classic spoked style; scuttling huntsman spiders the size of a human hand; tent-web spiders, which drape their vast pavilions of silk across gaps in the trees. Rogers calls the latter kind the "condo" web, because each one is akin to an apartment complex for creatures with eight legs – they contain hundreds of gleaming eyes, belonging to tens of individual spiders."You'll have many females at different levels in this massive web, and then many males hanging around and the edges," says Rogers. These communal webs are also a favourite of tiny Argyrodes spiders, who turn up to steal prey from, and occasionally eat, their much larger hosts. "On Guam these [condo webs] go from ground level, all the way up to the canopy – they can be everywhere," she says.Getty ImagesBefore brown tree snakes invaded, Guam's limestone forests were home to 12 native bird species (Credit: Getty Images)Most of the time, the whole forest looks like it's been draped in wisps of artificial spider's web for Halloween. "It's enough that when you're hiking, it's common for the person in front to pick up a spider stick and knock down the webs as they go," says Rogers, "otherwise you'll be covered in spider webs… I love it, but it is difficult to get through."Wherever there is a gap in the trees, the entire space will be filled with webs from hundreds of different spiders, all catching the light at different angles. These collective efforts can easily occupy a space the size of a room. "I had a friend once who ran into the middle of one and just twirled in a circle, making a mummy out of himself with this massive bit of web," says Rogers," he did that to freak out the people he was with".On another occasion, an assistant of Rogers volunteered to help in the field – only to get a couple of metres into the forest and change her mind. "She was like, 'No, I'm out'," says Rogers.Even without the spiders, which in the native Chamorro language are known as sånye'ye', Guam's limestone forests would be a strange and hostile place, like nowhere else.Overhead, towering breadfruit trees mingle with the Jurassic shapes of cycads, along with tangantangan (castor-oil) and spiky pandanus trees, forming a low canopy of jungle which is frequently ripped up by typhoons. On the ground, there is very little soil. Instead, plants grow directly out of limestone karst, forcing their roots into tiny fissures in the rock – the forest sits atop an ancient coral reef which has been pushed up over millions of years to form a plateau. Coral heads still litter the forest floor, and where the rock has eroded away, there are sinkholes and caves."I think the unique thing is, just like, it's really difficult to walk because – well, imagine walking on sharp rocks," says Rogers. When she takes new field technicians out for surveys, they need time to acclimatise to this rocky ground – she talks about people getting their "karst legs". "It's like getting your sea legs – being able to walk without having to focus on every step," she says.AlamyBanana spiders are abundant in Guam’s forests, where they build large webs that are 50% bigger than those found on neighbouring islands (Credit: Alamy)So, when Rogers decided to take a survey of the number of spiders in 2012, she knew it would be a challenge.There had always been rumours that Guam was particularly spidery – and that this might have something to do with the absence of birds, which ordinarily love to eat them. However, Rogers explains that the island's population of roughly 180,000 people rarely travel to the other Northern Mariana Islands – while they are all part of a self-governing commonwealth, only Guam is a US territory. As a result, there is little opportunity for comparison. And scientists had never actually checked.To find out exactly how many arachnids had taken over Guam, Rogers and colleagues set about completing transect surveys in the island's forests. For these, researchers carefully picked their way over the jagged coral reef underfoot while unspooling a roll of tape in a straight line. As they went. they carefully counted the spider's webs in their path which still contained a fanged occupant, if they fell within a metre of the line.What the scientists found was a population of arachno-spectacular proportions: during the wet season, there were 40 times more spiders in Guam's forests than there were on the nearby islands of Rota, Tinian and Saipan – while in the dry season, when spider populations in the region usually escalate, there were 2.3 times more spiders on Guam. The webs of banana spiders on Guam were also around 50% bigger.Getty ImagesHuntsman spiders don't build webs, and instead chase down their prey before injecting them with venom (Credit: Getty Images)All year round, Guam's forests were positively glistening with webs: the team found 1.8 and 2.6 webs per metre of transect line in the wet seasons and dry seasons respectively. If extrapolated to Guam's entire forest area this would amount to between 508 and 733 million spiders in total, scampering around on their webs and sucking the juices out of their victims. That's assuming one spider per web, though there are often far more – and only counting those living within two metres of the ground. In terms of spidery body parts, the forest ends up hosting at least 4,064,000,000 eyes and an equal number of hairy, articulated legs. Rota, Tinian and Saipan are free of brown tree snakes and still have healthy bird populations, so the study suggests that Guam's spider population might once have been unremarkable – before exploding over the last few decades, in the absence of birds. This is partly because of the avian predilection for creatures with eight legs, and also because they compete for the spiders' insect prey – and it fits with research conducted in the Bahamas, which has found that spiders are about 10 times more abundant on islands without lizards.Since the brown tree snake's arrival, the banana spider's existence on Guam has become so comfortable, they have even stopped adding "stabilimenta" to their webs. It's not known why spiders add these mysterious decorations, which usually involve zig-zag patterns made from opaque white threads. One idea is that they warn birds of the web's presence, preventing them from accidentally flying into them – and this is backed up by the feature's unusually low frequency on bird-less Guam. A stubborn invader Though brown tree snakes are thought to have been shifting the balance of Guam's ecosystems since they were introduced shortly after World War Two, this went largely undetected for at least four decades. By the late 1980s, ecologists had noticed that something was wiping out the island's birds – but no one had any idea what.Julie Savidge, who was a PhD student at the time, decided to track down this mystery killer, which was suspected to be either a pesticide or a virus. Her research – published in 1987 – revealed that it was, in fact, down to the snakes. As island species, most of Guam's birds had no evolutionary programming that could help them to avoid the reptiles' insatiable appetite for poultry – there are no native snakes, so their ancestors were equally clueless. Consequently, when the predators arrived they found an all-you-can-eat buffet of defenceless birds which quickly became dinner.By the time Savidge had worked out what was going on, for most avian species it was already too late. The Guam flycatcher was last seen in the wild in 1984, and this wide-eyed little ball of feathers is now considered extinct. Others have had a narrow escape.AlamyThe US Department of Agriculture uses sniffer dogs to intercept brown tree snakes at Guam's ports and airports and prevent them from escaping to other islands (Credit: Alamy)The lurid ginger and iridescent green Guam kingfisher was considered extinct in the wild until earlier this year, when nine captive individuals were introduced to the Palmyra Atoll, some 5,879 km (3,653 miles) away from their natural home. The Guam rail, a flightless bird known locally as the Ko'ko', was also once listed as extinct in the wild. Today their reddish-brown bodies and houndstooth bellies can be seen scurrying around Rota and the Cocos islands, where they have been introduced.But only in recent years have researchers begun to uncover the true scale of ecological chaos caused by brown tree snakes. The skinny reptiles are highly elusive, soundlessly slipping through Guam's forests and suburbs at night, tentatively tasting the air with their tongues to detect the scent of their next meal. But their slight appearance and the surprising rarity of sightings belies their wiles.As it turns out, brown tree snakes are not normal predators. There are few limits on what this ambitious eater will attempt to swallow, and they regularly consume animals that are 70% of their own body weight – equivalent to a 60kg (136lb) human eating a small red kangaroo in a single sitting.Recently, a team of scientists led by Rogers were monitoring the survival of fledgling Såli, a kind of forest starling that has managed to eke out an existence in the vicinity of Andersen Air Force Base, a US facility at the north of the island. The researchers attached radio monitors to the young birds and followed up to see what had happened to them. Often, the equipment was discovered inside the digestive systems of brown tree snakes. But there were also more grisly discoveries: dead birds that hadn't been eaten.To their surprise, they found that many fledglings were being killed by the snakes, then abandoned: when their corpses were tracked down, they were coated in a silvery residue of snake saliva, a situation the scientists informally described as "sliming". Around half the time, the snakes were killing birds that were too large for them to swallow.Ruth HufbauerHalde Rogers and her team found 40 times more spiders on Guam during the wet season than there were on neighbouring islands (Credit: Ruth Hufbauer)But brown tree snakes aren't just greedy – they're also supremely effective hunters, with acrobatic skills that mean they can find prey in even the most inaccessible spaces. To help keep the remaining Såli safe from brown tree snakes, conservationists have been providing nest boxes and fortifying them with metal "baffles" – slippery metal poles, 3ft (0.9m) long and around 15cm (6in) across, which no snakes are supposed to be able to climb. Alas, they had been unaware of the crafty predator's special talent.In 2021, a research team led by Savidge – now a professor in the department of fish, wildlife and conservation biology at Colorado State University in the US – discovered "lasso climbing", a method entirely new to science."Brown tree snakes can literally wrap themselves around a cylindrical baffle, hook their tail around their head and then shimmy up like a human climbing up a coconut tree," says Pollock, who describes the research as "mind blowing". An irreversible shift In the last few decades, conservationists and wildlife officials have used every conceivable method in an attempt to eliminate the brown tree snake from Guam – but the reptiles are winning. There have been visual searches, repellents, irritating substances, traps, poisons and deadly chemicals.Researchers have even searched for viruses that could be used as bioweapons against brown tree snakes, to eliminate large numbers without affecting other wildlife. This method would operate something like myxomatosis in rabbits, which has been widely circulated on purpose – including in France, illegally, and Australia – with the intention of reducing their numbers. It has also caused widespread suffering, however.But despite intense efforts, and an annual budget for snake control measures that's now around $3.8m (£2.9m), the invaders have proved to be impossible to remove in large numbers. That is, except from a few tiny patches of land. Take the Habitat Management Unit at Andersen Air Force Base on Guam. It just so happens that regular over-the-counter acetaminophen (paracetamol) is particularly toxic to brown tree snakes, with even the largest individuals perishing after an 80mg dose – around a sixth of the amount found in one standard 500mg tablet for humans. After a comprehensive programme which involved baiting them with poisoned food – along with erecting a "snake-proof" fence around the entire area to prevent it being immediately recolonised – their numbers dropped significantly on the air base.Alas, many scientists believe that it would be impossible to remove significant numbers of brown tree snake in a similar way from Guam's forests, let alone altogether. This is despite there being some urgency to do so, because the very forest itself is at risk.AlamyWildlife officials regularly trap brown tree snakes in an attempt to reduce their numbers in certain parts of Guam (Credit: Alamy)It's thought that roughly 70% of Guam's native trees relied on birds to disperse their seeds. But in the disconcertingly silent forest landscape today, many trees drop their fruit directly on the ground – only for it to rot where it fell. Some seeds won’t germinate at all unless the flesh has been eaten, says Rogers, while others struggle to grow in the shadow of their parent tree. With each passing year that the island's avian fruit, nut and seed-eaters don't turn up, the tree species that they used to rely on are dying out.The forest is also starting to develop holes. In a healthy ecosystem, when a tree falls it creates a temporary gap – and this immediately becomes the site of intense competition, as different plants jostle to fill the space. "It's like if you take down a building in the middle of New York City, it's a prime real estate, there's going to be a lot of people wanting to put a building up there," says Rogers. But on Guam this doesn't happen. Without birds to disperse seeds, often there is nothing on the ground that can germinate, so regeneration is extremely slow. The structure of the forest is changing, and soon there may be no easy way back.For now, Guam's brown tree snakes and the army of spiders they have created are safe. And their reign may last for some time yet.--For essential climate news and hopeful developments to your inbox, sign up to the Future Earth newsletter, while The Essential List delivers a handpicked selection of features and insights twice a week.For more science, technology, environment and health stories from the BBC, follow us on Facebook and Xand Instagram. | 2024-11-08T04:17:45 | en | train |
42,025,167 | ddaniel100 | 2024-11-02T09:36:29 | null | null | null | 15 | null | [
42025319,
42025259,
42025367,
42025175
] | null | true | null | null | null | null | null | null | null | train |
42,025,174 | peutetre | 2024-11-02T09:38:59 | null | null | null | 2 | null | [
42025621
] | null | true | null | null | null | null | null | null | null | train |
42,025,190 | fanf2 | 2024-11-02T09:42:02 | TestResult: a Rust error type that panics early (2023) | null | https://www.bluxte.net/musings/2023/01/08/improving_failure_messages_rust_tests/ | 1 | 0 | [
42025617,
42025616
] | null | null | null | null | null | null | null | null | null | train |
42,025,197 | Tomte | 2024-11-02T09:43:35 | Read Literature in Translation | null | https://www.youtube.com/watch?v=v79fqaSiJQk | 2 | 0 | [
42025613
] | null | null | null | null | null | null | null | null | null | train |
42,025,200 | croes | 2024-11-02T09:43:57 | Project Rodeo, the Tesla effort pushing the limits of self-driving technology | null | https://www.businessinsider.nl/inside-project-rodeo-the-tesla-effort-pushing-the-limits-of-self-driving-technology/ | 3 | 1 | [
42025333
] | null | null | null | null | null | null | null | null | null | train |
42,025,217 | ahamez | 2024-11-02T09:46:49 | Cppfront v0.8.0 | null | https://github.com/hsutter/cppfront/releases/tag/v0.8.0 | 20 | 2 | [
42037088,
42025484,
42025408
] | null | null | null | null | null | null | null | null | null | train |
42,025,221 | leogout | 2024-11-02T09:47:48 | Show HN: RasperDucky, an Implementation of DuckyScript3 for Raspberry Pico | The past few weeks I've been working on an implementation of DuckyScript3 used for keystroke injections attacks. See [1] for the official RubberDucky USB key page.<p>My goal is to write a python interpreter to port as much commands as possible to run them on a RaspberryPi pico. It's a fun side project that I've loved working on lately.<p>If you have any questions, feel free to ask. As mentioned in the README of the project, I am in no way affiliated with Hack5, which is the creator of the original USB RubberDucky and the DuckyScript language.<p>Here is a link to another project on this topic, [2] which inspired me for doing this. At first I tried to contribute to this project but I got no answers from the owner and the way the code was written didn't correspond to the way I would have done it, so it gave me the motivation to create my own.<p>I've learned lots and lots thanks to crafting interpreters [3], written by Bob Nystrom<p>[1]: <a href="https://shop.hak5.org/products/usb-rubber-ducky" rel="nofollow">https://shop.hak5.org/products/usb-rubber-ducky</a>
[2]: <a href="https://github.com/dbisu/pico-ducky">https://github.com/dbisu/pico-ducky</a>
[3]: <a href="https://github.com/munificent/craftinginterpreters">https://github.com/munificent/craftinginterpreters</a> | https://github.com/leogout/rasper-ducky | 2 | 0 | null | null | null | no_error | GitHub - leogout/rasper-ducky: A DuckyScript 3 implementation targeting Raspberry pico | null | leogout |
RasperDucky
An open-source implementation of DuckyScript3 with CircuitPython for Raspberry Pi Pico
This project aims to provide an open-source implementation of DuckyScript3, a scripting language used for automating tasks on USB Rubber Ducky devices. Its goal is to achieve a 1-to-1 implementation of DuckyScript3 targeting a Raspberry Pi Pico and Raspberry Pi Pico 2.
Getting Started
Installation on Raspberry Pi Pico 2
Install CircuitPython 9.2.0
Go to the CircuitPython website and download the CircuitPython 9.2.0 firmware for Raspberry Pi Pico 2 using the "DOWNLOAD .UF2 NOW" button.
Hold the BOOTSEL button on your Raspberry Pi Pico 2 and connect it to your computer using a USB cable. Once RPI-RP2 appears in the device manager, release the BOOTSEL button.
Open your RPI-RP2 drive and drag and drop the downloaded .UF2 file to the Raspberry Pi Pico 2.
Once the Raspberry Pi Pico is restarted, it will appear as CIRCUITPY in the device manager.
Install the rasper-ducky library
Connect your Raspberry Pi Pico to your computer and copy/paste the content of the rasper_ducky folder to the CIRCUITPY drive.
Edit the payload.dd file at the root of the CIRCUITPY drive to change the payload.
You can try it out with this simple script, which will open a PowerShell window and print "Hello, World!" 3 times, separated by a space:
DEFINE #COUNT 3
FUNCTION open_powershell()
GUI R
STRINGLN powershell
END_FUNCTION
FUNCTION hello_world()
$x = 0
WHILE ($x < #COUNT)
DELAY 500
STRING Hello, World!
SPACE
$x = $x + 1
END_WHILE
END_FUNCTION
open_powershell()
DELAY 1000
hello_world()
Debugging
To debug the script, connect to the Raspberry Pi Pico 2 using Putty or similar and use the serial console. I've seen ports up to COM8 on my computer so try them all until you find the correct one.
Once connected, you should see the output of the script in the serial console.
Disclaimer
I am not affiliated with Hak5 or USB Rubber Ducky in any way. This is a side project and I do it for fun.
There are some choices in the DuckyScript3 language that I dislike and may want to change in the future, maybe in a fork of this project. For example, in DuckyScript3 :
Functions do not take parameters and can't return values
Every variable is a global variable, no scoping is implemented, variables declared in functions will bleed into the global scope
No mathematical operator precedence, requiring parenthesis for simple expressions (10 + 2 * 3 will be interpreted as (10 + 2) * 3)
Variables can only be integers
DELAY, STRING and STRINGLN don't accept variables
No for loops, only WHILE loops are supported
Need to declare variables before using them
No BREAK or CONTINUE statements in loops
Features
Raspberry Pi Pico Support: Execute scripts on a Raspberry Pi Pico.
Full test suite: Ensure the implementation is correct and reliable.
Open Source: Contribute to the project and help me improve the implementation, please refer to the CONTRIBUTING.md file.
Partial DuckyScript3 Syntax Support: See INSTRUCTIONS.md for the supported syntax.
Additional features (not included in DuckyScript3 but nice to have):
Keyboard layouts: RD_KBD WIN FR for French keyboard on Windows, RD_KBD MAC UK for UK keyboard on Mac, etc.
Operators precedence: 10 + 2 * 3 / 4 will be interpreted as 10 + ((2 * 3) / 4)
No need for VAR: No need to declare variables before using them, just use them directly : $x = 10 is equivalent to VAR $x = 10
Random character from a string: RANDOM_CHAR_FROM aAzZ!#1,;:!() will type a random character from the given string
Roadmap
Complete the roadmap
Derecursify parts of the parser to limit stack overflows
Complete 1-to-1 DuckyScript3 implementation
WAIT_FOR_BUTTON_PRESS
BUTTON_DEF
LED_R / LED_G (only one green led on the Raspberry Pi Pico)
RANDOM_LOWERCASE_LETTER, RANDOM_UPPERCASE_LETTER, RANDOM_LETTER, RANDOM_NUMBER, RANDOM_SPECIAL, RANDOM_CHAR
HOLD / RELEASE
RESTART_PAYLOAD / STOP_PAYLOAD
Improve error handling and debugging features
Custom backward compatible features
Pass a variable to DELAY
Pass a variable to STRING and STRINGLN
Add support for # comments
Prevent the user from shadowing a keyword with a DEFINE
Add a CONTINUE and a BREAK statement to loops
Add a PRINT statement to debug scripts
Add a RANDOM_CHAR_FROM instruction to generate random characters from a given string
Custom backward incompatible features
Scope variables
Thanks
Thanks to @hak5 for the original DuckyScript language.
Thanks to @munificent for the crafting interpreters web book, which was a great help to understand how to implement a scripting language.
Thanks to @dbisu for Pico-Ducky, which was the starting point for this project.
Thanks to the Team behind CircuitPython for the amazing work on this firmware and their support.
Thanks to the rest of the community for making this fun little project possible (AdaFruit, MicroPython, etc.).
| 2024-11-08T02:40:15 | en | train |
42,025,228 | todsacerdoti | 2024-11-02T09:49:13 | Why Pg_dump Is Amazing | null | http://rhaas.blogspot.com/2024/11/why-pgdump-is-amazing.html | 2 | 0 | [
42025597
] | null | null | null | null | null | null | null | null | null | train |
42,025,233 | dnetesn | 2024-11-02T09:50:53 | Cat Is Listening to You | null | https://nautil.us/your-cat-is-listening-to-you-1045745/ | 3 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,235 | timbilt | 2024-11-02T09:51:03 | World's oldest tree? Genetic analysis traces evolution of iconic Pando forest | null | https://www.nature.com/articles/d41586-024-03570-4 | 3 | 0 | null | null | null | no_error | The world’s oldest tree? Genetic analysis traces evolution of iconic Pando forest | null | Kudiabor, Helena |
DNA samples from one of the world’s largest and oldest plants — a quaking aspen tree (Populus tremuloides) in Utah called Pando — have helped researchers to determine its age and revealed clues about its evolutionary history.By sequencing hundreds of samples from the tree, researchers confirmed that Pando is between 16,000 and 80,000 years old, verifying previous suggestions that it is among the oldest organisms on Earth. They were also able to track patterns of genetic variation spread throughout the tree that offer clues about how it has adapted and evolved over the course of its lifetime. The findings were posted on the bioRxiv preprint server on 24 October1. The work has not yet been peer reviewed.“It’s just pretty cool to study such an iconic organism,” says co-author Rozenn Pineau, a plant evolutionary geneticist at the University of Chicago in Illinois. “I think it’s important to draw people’s attention on natural wonders of the world.”One very big treePando — whose name means ‘I spread’ in Latin — consists of some 47,000 stems that cover an area of 42.6 hectares in Utah’s Fishlake National Forest. Because of the way the plant reproduces, this collection of aspens is technically all one tree, supported by a single, vast root system. Pando is triploid, meaning that its cells contain three copies of each chromosome, rather than two. As a result, Pando cannot reproduce sexually and mix its DNA with that of other trees, and instead creates clones of itself.The trees’ lessons: climate records are written in tree ringsAlthough this process generates offspring that are genetically identical, they can still accumulate genetic mutations as their cells divide. Biologists are interested in these variations because they provide information on how the plant has changed since the first seedling sprouted. Some studies have explored the spread of new mutations in plants and fungi that reproduce clonally, but few have investigated centuries-old plants like Pando.“It’s kind of shocking to me that there hasn’t been a lot of genetic interest in Pando already, given how cool it is,” says study co-author William Ratcliff, an evolutionary biologist at the Georgia Institute of Technology in Atlanta.The researchers collected samples of roots, bark, leaves and branches from across the Pando clone, as well as from other, unrelated quaking aspen trees for comparison. They extracted DNA from the samples, then sequenced and analysed a subsection of the genome.After removing variants that were found in both Pando and neighbouring trees, as well as mutations found in just one sample, the researchers were able to review nearly 4,000 genetic variants that had arisen as Pando cloned itself repeatedly over millennia.Analysing the patterns of these mutations revealed surprising results. “You would expect that the trees that are spatially close are also closer genetically,” says Pineau. “But this is not exactly what we find. We found a spatial signal, but that is much weaker than what we expected.” Physically close trees did share more similar mutations than those that were far apart — but only slightly more. However, over a smaller scale of 1–15 metres the trend was stronger, with stems that were closer together having significantly more shared mutations. Pando covers an area of more than 40 hectares, “but it almost looks like it’s a well-mixed pot of genetic information”, says Ratcliff.Protective mechanismBy inputting Pando’s genetic data into a theoretical model that plots an organism’s evolutionary lineage, the researchers also estimated Pando’s age. They put this at between 16,000 and 80,000 years. “It makes the Roman Empire seem like just a young, recent thing,” says Ratcliff.The team also considered reasons for the tree’s remarkable endurance. Pineau says that Pando being triploid might lead to “bigger cells, bigger organisms, better fitness”, and that existing clones might be more durable than new mixed offspring.Philippe Reymond, who researches interactions between plants and herbivores at the University of Lausanne in Switzerland, says that the findings hint that “plants and trees have a mechanism to protect the genome” from the accumulation of harmful genetic mutations, a suggestion that is “quite interesting for many scientists”. He adds that future studies could search for this exact mechanism at the cellular level.Ratcliff is also keen for more studies to be done on Pando’s genetic history. “I would love to make a call for people to work on these kinds of organisms,” he says.
| 2024-11-08T13:30:23 | en | train |
42,025,240 | dnetesn | 2024-11-02T09:51:44 | Plants, Animal Behavior and the Origins of Medicine | null | https://worldsensorium.com/the-healing-instinct-plants-animal-behavior-and-the-origins-of-medicine/ | 1 | 0 | null | null | null | no_error | The Healing Instinct: Plants, Animal Behavior and the Origins of Medicine | 2024-10-31T20:40:34+00:00 | null |
Photograph by Gayil NallsBy Gayil NallsSign up for our monthly newsletter!L ast month, I observed a fascinating interaction between my dog and a neighbor’s aging companion, Juno. My dog, who often has his nose in a plant, was grazing on specific grass species, seemingly in an attempt to soothe his intestinal discomfort. Later that morning, I watched Juno, a white-faced senior dog, delicately eat a small amount of yarrow (Achillea millefolium) while my dog curiously observed. My dog then sniffed and licked the flower but ultimately chose not to eat any.
Yarrow, well-known to humans for its medicinal properties, has been used for centuries as a digestive aid and mild astringent, often to treat diarrhea. The plant contains compounds like flavonoids and sesquiterpene lactones, which help reduce inflammation. However, veterinary medicine suggests that the same sesquiterpene lactones can sometimes exacerbate gastrointestinal issues in dogs and may even cause neurological symptoms, particularly in those sensitive to the plant.
The observation of animals interacting with medicinal plants is not uncommon. Cats, dogs, wolves, and even bears have been known to consume specific grasses and plants when experiencing digestive distress or infections. These plants are thought to act as purgatives or anti-inflammatories. This instinctive behavior is known as zoopharmacognosy—the act of animals self-medicating through natural substances found in their environment.
Zoopharmacognosy is an extraordinary phenomenon seen across the animal kingdom, illustrating that animals possess an innate ability to treat their own ailments. One particularly striking example comes from recent research on an orangutan in Sumatra. This individual, named Rakus, was observed creating a poultice from medicinal plants to treat an open wound on his face. The observation, published in Scientific Reports and Nature, marked the first scientifically recorded instance of a wild animal using known medicinal plants to heal itself in this way. Rakus chewed specific leaves without ingesting them and used his fingers to smear the resulting juice over his wound for about seven minutes. Remarkably, he repeated the treatment the following day, and just over a week later, the wound had fully healed.
This behavior of self-medication is also seen in many other species. For example, Canadian geese are known to consume particular leaves to expel tapeworms, while dusty-footed woodrats line their nests with aromatic plants such as gnawed bay leaves that act as fumigants, keeping parasites at bay. In Kenya, pregnant elephants have been observed eating the leaves of Boraginaceae plants to induce labor, likely due to the uterine-stimulating properties of the plant’s chemicals.
Such examples highlight that animals rely on a combination of instinct, learned experience, and environmental cues to identify plants that can aid in their healing. This behavior underscores a deep, evolutionary connection between animals and their natural environments, revealing sophisticated methods of health regulation that extend far beyond basic survival.
The healing instinct observed in animals has deep roots in the natural world, and it is believed to have significantly influenced early human knowledge of herbal medicine. Our ancestors, living closely with nature, likely noticed how animals would seek out specific plants to alleviate their ailments. This innate behavior, passed down through generations of animal species, served as a form of natural wisdom for humans to tap into. Indigenous cultures across the globe have long relied on such observations to guide their use of plants in treating a wicncnde range of disorders. The behavior of animals, particularly in their selection of medicinal herbs, became an early blueprint for developing remedies that would later evolve into the sophisticated practices of herbalism and modern pharmacology. This close connection between animal behavior and human medicine underscores a shared biological intelligence—one rooted in survival and the healing power of nature. Today, zoopharmacognosy continues to inspire scientific research, reaffirming that the animal kingdom holds untapped knowledge of medicinal plants that may benefit both human and veterinary medicine.
This intersection of nature and science continues to offer insights that could lead to the discovery of novel therapeutic approaches, shaped by the power of plants and the wisdom of the animal world.
Dogs sometimes eat grass as part of a natural behavior, often for self-medicating or digestive relief. While many grass species are safe, some are better suited for dogs to nibble on without causing harm. Here are grasses that are both safe and beneficial for dogs:
Wheatgrass (Triticum aestivum) is rich in vitamins, minerals, and antioxidants, wheatgrass is often used in pet supplements. It can aid digestion, improve metabolism, and provide a nutritional boost.
Barley Grass (Hordeum vulgare) is a nutrient-dense with high fiber content, barley grass helps support digestion and is safe for dogs to eat in moderation. It may assist with stomach upset or constipation.
Oat Grass (Avena sativa) is easy to digest and contains vitamins A, B, and C, making it a healthy choice for dogs. It can soothe digestive issues and is gentle on the stomach.
Lemongrass (Cymbopogon citratus) is known for its calming effects, lemongrass can help with digestive problems. However, it should be offered in small amounts, as large quantities might cause stomach upset.
Bamboo Grass (Phyllostachys spp.) is non-toxic to dogs and can be eaten safely. Dogs sometimes chew on bamboo leaves, which contain some fiber and may help with digestion.
Rye Grass (Lolium perenne) is not nutrient-rich like wheatgrass or barley grass, but is safe for dogs to nibble on when they are trying to soothe an upset stomach.
Fescue Grass (Festuca spp.) is safe and non-toxic for dogs, fescue is commonly found in lawns and can be eaten without harm when dogs graze on it.
Gayil Nalls, Ph.D., is the creator of World Sensorium and founder of the World Sensorium/Conservancy.
References
Vaidyanathan, Gayathri, Nature, Vol 629, 23 May 2024, p.737
Greene, Alexander, et al. Asian elephant self-medication as a source of ethnoveterinary knowledge among Karen mahouts in northern Thailand. https://pubmed.ncbi.nlm.nih.gov/32387460/
The Proposed Mechanism of Leaf-swallowing and its Effect on Infection https://www.researchgate.net/figure/The-proposed-mechanism-of-leaf-swallowing-and-its-effect-on-infections-with_fig2_225938467
Krieger, Marilyn. Dusty-Footed Woodrats Help Save a Canyon
Plantings Print Annual 2024
Do you have the 2024 Plantings print annual?
Plantings cultivates innovative ideas and fresh perspectives, nurturing the global conservation community. Our readers find inspiration in forward-thinking individuals and approaches dedicated to fostering a better life for the planet and all its inhabitants.
The 2024 edition of Plantings is available in our store for shipping.
| 2024-11-08T04:47:25 | en | train |
42,025,242 | sesigl89 | 2024-11-02T09:52:15 | null | null | null | 1 | null | [
42025243
] | null | true | null | null | null | null | null | null | null | train |
42,025,245 | timbilt | 2024-11-02T09:53:02 | New record-breaking Prime Number, 41M Digits Long | null | https://www.scientificamerican.com/article/new-prime-number-41-million-digits-long-breaks-math-records/ | 2 | 1 | [
42025348,
42025588,
42025586
] | null | null | no_error | New Prime Number, 41 Million Digits Long, Breaks Math Records | 2024-11-01T15:30:00+00:00 | Jack Murtagh | Thousands of computers across the world are currently scouring the number line in a scavenger hunt for rare mathematical gems. Prime number enthusiasts, looking for larger and larger numbers that are divisible only by 1 and themselves, muster vast amounts of computing power and algorithmic ingenuity in hopes of etching their name into the scrolls of math history.The latest entry comes from Luke Durant, a researcher in San Jose, Calif. Durant’s discovery overturned the former record holder, which sat uncontested for nearly six years, an unprecedentedly long reign in the modern search for ever larger prime numbers. The gap makes sense: as primes grow, they spread further apart, making each new find harder than the last.The new prime contains a mind-boggling 41,024,320 digits. To put that in perspective, the estimated number of atoms in the observable universe clocks in at around 80 digits. Every additional digit increases a number by 10 times, so the size of the new prime lives far beyond human intelligibility. Primes play a major role in pure math, where they’re main characters in a field called number theory, and in practice, where, for example, they underlie widely used encryption algorithms. A prime with 41 million digits won’t immediately join the ranks of useful numbers, but for now, it adds a feather in the cap of a community that longs to apprehend the colossal.On supporting science journalismIf you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.Durant’s success stems in part from new clever software from the Great Internet Mersenne Prime Search and in part from heavy-duty hardware and computational muscle that he personally mobilized for the pursuit. By assembling a “cloud supercomputer” spanning 17 countries, he ended a long tradition of personal computers discovering primes.Prime numbers are often called the “building blocks of math” because every whole number greater than 1 has a fingerprint as the product of a unique collection of primes. For example, 15 is the product of the primes 5 and 3, whereas 13 cannot be subdivided like this because 13 is prime. The study of these numbers dates back at least to the ancient Greeks. In 300 B.C.E. Euclid proved in his textbook Elements that infinitely many primes exist, and mathematicians, both professional and amateur, have relished the hunt for them ever since.While the first string of primes—2, 3, 5, 7, 11, and so on—is easy to find, the task gets considerably more challenging as the numbers get larger. For millennia, people found primes by hand—until 1951, when computers took over the search. But even silicon bounty hunters struggle to spot primes in the far reaches of the number line because testing the primality of an enormous number is nontrivial. To cope, researchers deploy every little optimization trick they can to speed up their tests or narrow their hunting ground, thereby boosting their chances of finding a new prime.Consider the number 99,400,891. How would you determine whether or not it’s prime? You could simply divide it by every smaller number and check if it has any divisors (in addition to 1 and itself). But that’s nearly 100 million cases to check for a relatively puny eight-digit number. You would save significant work by realizing that you don’t need to check every number up to the target, just the prime numbers. Why? You only need to find one divisor (one number that cleanly divides 99,400,891 with no remainder). We know that any nonprime divisor could be further broken down into its prime factors—if your target is divisible by 15, then it’s also divisible by the primes 5 and 3, so you only need to check the latter to determine primality. Further savings would come from the insight that you don’t need to check every smaller prime either, only those up to the square root of 99,400,891 (the number that when multiplied by itself gives you this eight-digit result). If none of those smaller primes divide it cleanly, then you can stop looking because the product of any two numbers larger than the square root of 99,400,891 will exceed it. These efficiency tricks slash our search drastically, from around 100 million numbers to only 1,228 (the number of primes less than the square root of 99,400,891). For those curious, 99,400,891 = 9,967 × 9,973, so it’s not prime.Those shortcuts did wonders for an eight-digit number, but how did Durant reach 41,024,320 digits? To graduate the search from the merely massive to the truly gargantuan, he and many other seekers focus on particular types of prime numbers. Mersenne primes, named for Marin Mersenne, the French theologian who studied them in the 17th century, take a special form. You get them by multiplying 2 by itself some number of times and then subtracting 1, as described in the equation 2n – 1. Mersenne noticed that when you plug in different values for n, you sometimes get a prime number. Specifically, 2n – 1 can only yield a prime when n itself is prime, and even then it’s not guaranteed. What makes Mersenne primes special from a prime hunter’s perspective is that we know a fast method (called the Lucas-Lehmer primality test) for checking whether numbers of the form 2n – 1 are prime. That test is much faster than any known general methods for numbers without that special form.The Lucas-Lehmer test fuels the Great Internet Mersenne Prime Search project, which launched in 1996 and enables any volunteer to download a free code that searches for Mersenne primes to run on their computers. The crowdsourced approach and the focus on Mersenne primes have proved successful. The seven largest known primes are all Mersenne primes and were all found by participants of the project. Note that smaller unknown primes certainly exist, but because we don’t know efficient methods for checking them, they’ll remain in the shadows for now.All told, project volunteers have found 18 new Mersenne primes, 17 of which owe their discovery to the personal computers of hobbyists. Durant, a 36-year-old former Nvidia engineer, just broke that streak. Nvidia, which recently briefly overtook Apple as the world’s most valuable company, designs specialty computer chips called graphics processing units (GPUs). As the name suggests, GPUs were originally invented to accelerate the rendering of graphics, but they also excel at other tasks involving highly parallelized computation, in which many processors run simultaneously to solve problems. Those problems include training neural networks such as GPT-4, mining cryptocurrency and, as it turns out, foraging for primes. Durant assembled a global supercomputer by buying processing time from various cloud GPU providers. At its peak, Durant’s project churned through about 12 times as many numbers as every other computer involved in the Mersenne prime search combined.In addition to the heavy-duty hardware, the Mersenne prime search software also got a notable upgrade since the last discovery. It replaced the superfast Lucas-Lehmer test for certifying Mersenne primes with a super-duper-fast probable prime test. Given a number to check, the latter test either confirms that it’s not prime or says that it’s probably prime. Probable primes have a very small chance of turning out to be nonprime. Only once a computer finds a probable prime do Mersenne prime search volunteers run the full-fledged Lucas-Lehmer test to remove any doubt. Durant’s new prime passed the probable prime test on October 11. Then, on October 19, a year after Durant started searching, independent tests by the Mersenne prime search confirmed that he had indeed found a needle in a haystack: 2136,279,841 – 1 is the largest known prime number.It exceeds the previous record holder by more than 16 million digits. If that didn’t earn enough glory, Durant has also unearthed the largest known “perfect number.” A perfect number equals the sum of its divisors (excluding itself); 6 is perfect because it’s divisible by 1, 2, 3 and equals the sum of 1 + 2 + 3. The second perfect number is 28. The 18th-century mathematician Leonhard Euler proved that every even perfect number can be generated from a Mersenne prime, so finding one promises a two-for-one deal on math discoveries.The well could dry up any time, though. We don’t know whether an infinite number of Mersenne primes (and therefore even perfect numbers) exist. Curiously, we don’t know whether any odd perfect numbers exist, a question that some call the oldest unsolved math problem.When asked how much money his project cost in an interview with Numberphile on YouTube, Durant said, “I believe it’s under $2 million.” That’s a hefty investment compared with the prime search project prize money of $3,000, which he plans to donate to the high school he attended, the Alabama School of Mathematics and Science. At this point, you might wonder why so many people spend their time and money trolling for primes that don’t have obvious real-world applications. In part, their efforts celebrate human curiosity and serve as a benchmark for our progress in mathematical computation. But I think the founder of the Mersenne prime search project, George Woltman, when asked this question in a Numberphile video, said it best: “It’s fun.” | 2024-11-08T14:13:00 | en | train |
42,025,246 | zaplin | 2024-11-02T09:53:06 | I applied to my dream job and got an instant response | null | https://gkampiotis.notion.site/What-is-this-115cf06f7b628087b38eda2735bed7dd | 2 | 1 | [
42025247,
42025582
] | null | null | null | null | null | null | null | null | null | train |
42,025,256 | timbilt | 2024-11-02T09:54:54 | The International Space Station Has Been Leaking for Five Years | null | https://www.scientificamerican.com/article/the-international-space-station-has-been-leaking-for-five-years/ | 18 | 9 | [
42026087,
42025826
] | null | null | null | null | null | null | null | null | null | train |
42,025,277 | beardyw | 2024-11-02T10:01:48 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,308 | albertkag | 2024-11-02T10:08:12 | Show HN: I Created Keyword Rank Checker for Site Linked to Google Search Console | null | https://chromewebstore.google.com/detail/keyword-rank-checker/nmldnjcblcihmegipecakhmnieiofmgl | 3 | 3 | [
42026511
] | null | null | null | null | null | null | null | null | null | train |
42,025,310 | moviexme | 2024-11-02T10:09:05 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,337 | eashish93 | 2024-11-02T10:15:19 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,349 | funorawa | 2024-11-02T10:18:40 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,360 | marban | 2024-11-02T10:21:19 | Digital ad market booms for Big Tech | null | https://www.axios.com/2024/10/31/digital-ad-market-boom-big-tech | 3 | 1 | [
42025603,
42025600
] | null | null | null | null | null | null | null | null | null | train |
42,025,372 | pxc | 2024-11-02T10:24:11 | GNU Emacs comes with a book-length introduction to computer programming | null | https://www.gnu.org/software/emacs/manual/html_node/eintr/index.html | 9 | 6 | [
42025410,
42050837,
42025572,
42025441
] | null | null | no_error | gnu.org | null | About the Author |
An Introduction to Programming in Emacs Lisp
The GNU Emacs website is at
https://www.gnu.org/software/emacs/.
To view this manual in other formats, click
here.
This is An Introduction to Programming in Emacs Lisp, for
people who are not programmers.
Distributed with Emacs version 29.2.
Copyright © 1990–1995, 1997, 2001–2024 Free Software
Foundation, Inc.
Printed copies available from https://shop.fsf.org/. Published by:
GNU Press, https://www.fsf.org/licensing/gnu-press/
a division of the email: [email protected]
Free Software Foundation, Inc. Tel: +1 (617) 542-5942
51 Franklin Street, Fifth Floor Fax: +1 (617) 542-2652
Boston, MA 02110-1301 USA
ISBN 1-882114-43-4
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.3 or
any later version published by the Free Software Foundation; there
being no Invariant Section, with the Front-Cover Texts being “A GNU
Manual”, and with the Back-Cover Texts as in (a) below. A copy of
the license is included in the section entitled “GNU Free
Documentation License”.
(a) The FSF’s Back-Cover Text is: “You have the freedom to
copy and modify this GNU manual. Buying copies from the FSF
supports it in developing GNU and promoting software freedom.”
This master menu first lists each chapter and index; then it lists
every node in every chapter.
| 2024-11-08T10:18:53 | en | train |
42,025,378 | gfortaine | 2024-11-02T10:25:10 | 2 hidden ways to get more from your Gmail address | null | https://gmail.googleblog.com/2008/03/2-hidden-ways-to-get-more-from-your.html | 5 | 2 | [
42030255,
42027594,
42025567,
42025392
] | null | null | null | null | null | null | null | null | null | train |
42,025,380 | mindfulrings | 2024-11-02T10:25:35 | null | null | null | 1 | null | [
42025381
] | null | true | null | null | null | null | null | null | null | train |
42,025,418 | aussieguy1234 | 2024-11-02T10:35:29 | AI Will Understand Humans Better Than Humans Do | null | https://www.wired.com/story/plaintext-ai-will-understand-humans-better-than-humans-do/ | 4 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,465 | lifeisstillgood | 2024-11-02T10:44:04 | Portsmouth City Council among UK authorities hit by cyber attack | null | https://www.bbc.co.uk/news/articles/cly2jyvx55do | 1 | 1 | [
42025490
] | null | null | null | null | null | null | null | null | null | train |
42,025,466 | xcodevn | 2024-11-02T10:44:28 | PodcastLM: An open-source AI podcast creator | null | https://github.com/podcast-lm/the-ai-podcast | 3 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,469 | Spawek | 2024-11-02T10:44:42 | SpawELO – small free matchmaking system for LAN parties | null | https://blog.spawek.com/SpawELO | 181 | 47 | [
42028077,
42032192,
42029974,
42025470,
42029126,
42030096
] | null | null | no_error | SpawELO: Matchmaking system for LAN parties | null | null |
Polska wersja
Hey! I'm Spawek, and this is the story of a project I built with my friends to make our LAN parties better.
LAN Party
For the past 16 years, we’ve been organizing at least one LAN party every year with a group of friends. These parties typically last 4-5 days, with around 12 people joining at peak times. It’s not just about gaming - we also enjoy partying, intense discussions (we’re all big nerds), board games, card games like Star Realms, and Mafia. Some people arive later, some leave earlier. Some go to take care of their kids for a while. The LAN party is a place of fun and chill.
It's a LAN party so, we play computer games a lot! Dota 2 is our main game, but we also play classics like Counter-Strike, Wolfenstein: Enemy Territory, Warcraft 3, Blooby Volley, and more.
Selecting teams - especially Dota 2 ones - was always hard for us. We decided to automate the team selection process. It led to a very interesting ideas and a neat solution. But to get there, you will need a bit of a context.
Dota 2
A single Dota 2 match lasts about 40 minutes and is usually played 5v5 (though 6v6 is possible). Unbalanced matches, like 4v5, tend to be one-sided. For us it's just more fun when everyone plays - so often games are unbalanced.
Better players use their skill advantage to level faster and earn more gold (a.k.a. snowball), which makes even small differences between players very noticible.
In our group, around half of us play Dota regularly, while others have only ever played during our LAN parties, so skill differences are huge. There are also some other factors:
How do we normally pick teams
To choose teams, we typically have two leaders (usually the best or least experienced players) who pick teammates in alternating turns, much like picking football teams in school.
There’s a twist: the first leader picks one player, the second picks two, then the first leader picks two, and so on, with a single-player choice as each leader’s final pick. This balances the advantage of picking first.
Example with 11 people:
Leaders:
- Spawek (me - leader 1)
- Bixkog (my brother - leader 2)
1st pick - team 1 - 1st pick = 1 player:
- Muhah
2nd pick - team 2 - regular pick = 2 players:
- Alpinus
- Dragon
3rd pick - team 1:
- Vifon
- Goovie
4th pick - team 2:
- Status
- Bania
5th pick - team 1 - last pick = 1 player:
- Hypys
6th pick - team 2 - last pick = 1 player:
- J
team 1
team 2
Spawek
Bixkog
Muhah
Alpinus
Vifon
Dragon
Goovie
Status
Hypys
Bania
J
As the differences between players are big, this process tends to pick similar or even the same teams each time, which is just boring for us. It doesn't work well when the number of players is uneven.
Problems with manual team selection
Takes time and becomes annoying
Often results in similar teams
No one wants to be a leader
Imbalanced teams, especially with uneven player numbers
Solving it with code!
During the last LAN party, we got frustrated with the process and coded something quickly to automate it.
1st step: gather data
We found some historical data and put it to Colab (all data and code is there if you want to play yourself).
data = [
{
"winner": ["Dragon", "Spawek", "Status", "Vifon"],
"loser": ["Bixkog", "Alpinus", "Hypys", "J", "Goovie"]
},
{
"winner": ["Bixkog", "Alpinus", "Vifon", "J", "Goovie"],
"loser": ["Dragon", "Hypys", "Spawek", "Status"]
},
{
"winner": ["J", "Hypys", "Spawek", "Status", "Vifon"],
"loser": ["Bixkog", "Dragon", "Goovie", "Alpinus"]
},
# ... (35 games in total)
]
Elo rating
Elo rating is often used in games. Players starts at 1000 Elo points, earning or losing points with each game.
The probability of P1 (player 1) winning with P2 is defined as:
$$ P(P1, P2) = \frac{1}{1 + 10^{\frac{P2-P1}{400}}} $$
So only the Elo difference between players (D) is relevant:
$$ D = P1 - P2 $$
$$ P(D) = \frac{1}{1 + 10^{\frac{-D}{400}}} $$
The chance of winning is 50% if you have the same Elo as your opponent, 64% if you have 100 Elo more, and 76% if you have 200 Elo more:
Elo for unbalanced teams
To get the Elo of each team, let's simply sum the Elo of all team members:
$$ Elo(team) = \sum_{}^{team}Elo(player) $$
It's obviously problematic for unbalanced teams, but let's start simple.
Let's create the first solution by giving players 20 Elo for winning a match, and taking 20 Elo for losing a match:
from collections import defaultdict
elo = defaultdict(lambda: 1000.0)
for game in data:
for winner in game["winner"]:
elo[winner] += 20.0
for loser in game["loser"]:
elo[loser] -= 20.0
We get some results:
Spawek: 1260
Bixkog: 1140
Dragon: 1100
Bania: 1020
Muhah: 1020
Vifon: 1000
Hypys: 980
Alpinus: 960
Status: 900
Goovie: 900
J: 860
With this, we could generate balanced teams by finding combinations with the smallest Elo difference:
from itertools import combinations
requested_players = ["Hypys", "Spawek", "Bixkog", "Muhah", "J", "Vifon", "Bania", "Goovie"]
best_diff = float('inf')
best_team = None
expected_elo = sum([elo[player] for player in requested_players]) / 2
for team1 in combinations(requested_players, len(requested_players) // 2):
team1 = list(team1)
team1_elo = sum([elo[player] for player in team1])
diff = abs(team1_elo - expected_elo)
if diff < best_diff:
best_diff = diff
team2 = [player for player in requested_players if player not in team1]
team2_elo = sum([elo[player] for player in team2])
best_teams = (team1, team2, team1_elo, team2_elo)
print(f"team 1: {best_teams[0]}, total Elo: {round(best_teams[2])}")
print(f"team 2: {best_teams[1]}, total Elo: {round(best_teams[3])}")
Results:
team 1: ['Hypys', 'Spawek', 'J', 'Vifon'], total Elo: 4100
team 2: ['Bixkog', 'Muhah', 'Bania', 'Goovie'], total Elo: 4080
Woohoo! We've got the first solution!
Improvement 1: multiple passes through the data
We only have 35 historical games, so passing through them only once to compute Elo seems like a waste.
In the Elo system, you will get more than 20 points when you win against a player having a higher Elo, same way you will lose less than 20 points if you lose against a player having a higher Elo. For a winner:
$$ update = 40 * (1 - P(win)) $$
The losing player loses exactly the same amount of Elo, that winner gets. E.g.
Spawek (Elo: 1260) wins vs Goovie (Elo: 900):
Spawek += 4.47 Elo
Goovie -= 4.47 Elo
Status (Elo: 900) wins vs Dragon (Elo: 1100):
Status += 30.38 Elo
Dragon -= 30.38 Elo
In colab:
def win_probability(A, B):
elo_diff = A - B
return 1.0 / (1.0 + pow(10.0, (-elo_diff / 400.0)))
def elo_update(winner, loser):
prob = win_probability(winner, loser)
return 40.0 * (1 - win_probability(winner, loser))
As we are using team Elo instead of individual players, let's split the update between all players in the team evenly. Let's go through the data multiple times and see how the results change:
from collections import defaultdict
elo = defaultdict(lambda: 1000.0)
for iteration in range(0, 401):
for game in data:
winner_elo = sum(elo[player] for player in game["winner"])
loser_elo = sum(elo[player] for player in game["loser"])
update = elo_update(winner_elo, loser_elo)
for winner in game["winner"]:
elo[winner] += update / len(game["winner"])
for loser in game["loser"]:
elo[loser] -= update / len(game["loser"])
We coded this during the LAN party and it was a really awesome feeling to see that the computations converge! We used that algorithm for the rest of the party, adding new data after each game.
The solution was quite good, but we've seen it creating clearly unbalanced matches from time to time. In such cases, we just added the match with the assumed winner to the data as a "fake game" and generated new teams.
I knew that for the next LAN party, we could do better than that.
Improvement 2: make it more like an ML model
Key ideas:
Use Elo as a model to predict team victory chances
Store model information in each player’s Elo
Train it like an ML model
Let's stick to the Elo system we are using. The model will compute Elo for each player. We will compare SUM(Elo) of teams to compute the probability of winning.
Let's define a simple L2 loss function. Our model will try to minimize loss over all games:
$$ loss = (\text{real_chance} - \text{predicted_chance})^2 $$
def loss(data, elo):
loss = 0.0
for game in data:
winner_elo = sum(elo[player] for player in game["winner"])
loser_elo = sum(elo[player] for player in game["loser"])
computed_probability = win_probability(winner_elo, loser_elo)
real_probability = 1.0
loss += (real_probability - computed_probability)**2.0
return loss
So to compute total loss we have to go over all games and:
Add the elo of all players in the winning team and subtract the Elo of all players in the losing team
Compute winning probability
Compute L2 loss
We defined the model defined. Now we need to train it somehow.
Backpropagation: chain rule
In each step backpropagation will help us move the Elo of each player to minimize the loss. It basically computes (through derivatives) how changing the Elo of each player will impact the loss.
First, we need to do a forward pass (computations we did on the last diagram) to compute loss, then we will move backward. We don't really need the loss itself, but we need it's derivative:
$$ loss = (real - predicted) ^ 2 $$
$$ loss' = 2 * (real - predicted) $$
Then we multiply the result of that step by the derivative of the next step - win probability:
$$ P(D) = \frac{1}{1 + 10^{\frac{-D}{400}}} $$
I used Wolfram Alpha to compute derivative:
$$ P(D)' = \frac{log(10)}{400*(1 + 10^{D/400})} - \frac{log(10)}{400*(1 + 10^{D/400})^2}$$
In the last step, we just add/subtract the Elo of each player depending on their team:
from collections import defaultdict
import math
def backpropagation(data, elo):
derivative = defaultdict(lambda: 0.0)
for game in data:
winner_elo = sum(elo[player] for player in game["winner"])
loser_elo = sum(elo[player] for player in game["loser"])
elo_diff = winner_elo - loser_elo
computed_probability = win_probability(winner_elo, loser_elo)
real_probability = 1.0
final_derivative = 2.0 * (real_probability - computed_probability)
win_probability_derivative = final_derivative * (
-math.log(10) / (400.0 * (1.0 + 10.0**(elo_diff / 400.0))**2)
+ math.log(10.0) / (400.0 * (1.0 + 10.0**(elo_diff / 400.0))))
for player in game["winner"]:
derivative[player] += win_probability_derivative
for player in game["loser"]:
derivative[player] -= win_probability_derivative
return derivative
We can directly use the backpropagation function to optimize our model:
from collections import defaultdict
LEARNING_RATE = 10_000.0
ITERATIONS = 10001
elo = defaultdict(lambda: 1000.0)
for i in range(0, ITERATIONS):
derivative = backpropagation(data, elo)
for player in derivative:
elo[player] += derivative[player] * LEARNING_RATE
The model managed to minimize the loss:
But the resulting Elo doesn't converge:
A quick peek at the data shows the reason: the model loss is very small and the model is overfitting. It basically learned all the games by heart and does not generalize.
Game:
winner elo: 9591, winner team: ['Bania', 'J', 'Spawek', 'Bixkog', 'Dragon']
loser elo: 7485, loser team: ['Hypys', 'Status', 'Alpinus', 'Goovie', 'Vifon']
computed probability: 0.999994567526197
real probability: 1
Game:
winner elo: 9964, winner team: ['Spawek', 'Bixkog', 'Status', 'Goovie', 'Dragon']
loser elo: 9431, loser team: ['J', 'Hypys', 'Alpinus', 'Vifon', 'Muhah']
computed probability: 0.9555615730980401
real probability: 1
Game:
winner elo: 9990, winner team: ['Hypys', 'Spawek', 'Bixkog', 'Alpinus', 'Bania']
loser elo: 9130, loser team: ['Status', 'Dragon', 'Muhah', 'J', 'Vifon']
computed probability: 0.9929703085603608
real probability: 1
(32 more games)
We have to solve this problem. The reason for which we create that model is to be able to construct good teams, not encode the history!
Improvement 3: making historic results probabilistic
To fight overfitting we could go with typical ML solutions: we could add L1/L2 regularization or add some noise to the input data.
But I have another idea! We should remember that the historical games were mostly good - quite even Dota 2 games. Playing many games with same teams would probably lead to one team winning more often, but definitely not winning 100% of games.
I fetched more data from the history to asses which games were "even" and, which were "clearly one-sided".
I set "even" games to have a win probability of 75%, and "clearly one-sided" games to have a win probability of 95%:
{
"winner": ["Bixkog", "Status", "Alpinus", "Muhah", "Vifon"],
"loser": ["Hypys", "Spawek", "Goovie", "Dragon", "Bania", "J"],
"win_probability": 0.75, # even
},
{
"winner": ["Spawek", "Bixkog", "Alpinus", "J", "Vifon"],
"loser": ["Hypys", "Status", "Goovie", "Dragon", "Muhah", "Bania"],
"win_probability": 0.95, # one-sided
},
(...)
Using that in the code instead of 100% probability will make it way harder for the model to learn all the games by heart: the Elo difference needed for 75% win probability is ~200, the Elo difference needed for 100% probability is in the range [~500, infinity].
After updating the loss and backpropagation functions to use real_probability = game["win_probability"] instead of real_probability = 1 everything looks great.
Loss goes down quickly:
Players' Elo converge and are on reasonable levels:
We are good to go!
The new system is ready to predict victory chances, even with uneven team sizes. Here’s our first lineup for the LAN party starting in two weeks:
team 1 (Elo: 2660)
team 2 (Elo: 2655)
Spawek
Hypys
Bixkog
Muhah
Bania
J
Goovie
Vifon
Status
More updates to come!
| 2024-11-08T11:24:10 | en | train |
42,025,510 | perihelions | 2024-11-02T10:50:25 | Spain's PM orders 10k troops and police to Valencia | null | https://www.bbc.com/news/articles/ced9l7799w9o | 5 | 1 | [
42045193,
42025566
] | null | null | null | null | null | null | null | null | null | train |
42,025,512 | skar3 | 2024-11-02T10:50:51 | Contribute to Online Freedom with Snowflake | null | https://pietro.in/en/posts/snowflake/ | 3 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,559 | AstroCoder768 | 2024-11-02T10:59:07 | null | null | null | 1 | null | [
42025560
] | null | true | null | null | null | null | null | null | null | train |
42,025,579 | thezipster | 2024-11-02T11:02:08 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,598 | AntonZ234 | 2024-11-02T11:05:58 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,623 | elsewhen | 2024-11-02T11:10:41 | An Okta login bug bypassed checking passwords on some long usernames | null | https://www.theverge.com/2024/11/1/24285874/okta-52-character-login-password-authentication-bypass | 3 | 1 | [
42026985,
42026114
] | null | null | null | null | null | null | null | null | null | train |
42,025,641 | mikeirvine | 2024-11-02T11:14:46 | Show HN: Launched UdyamQR, Create Custom Webpages, QR Codes for Small Businesses | It helps to Create and update your business webpage effortlessly, showcasing your products and services to potential customers and Generate unique QR codes for your shop, cafe, or business that customers can scan to learn more about your offerings. | https://udyamqr.com/ | 1 | 1 | [
42032403
] | null | null | null | null | null | null | null | null | null | train |
42,025,668 | rustoo | 2024-11-02T11:20:01 | A Growth Mindset at Work Can Help in Unexpected Ways | null | https://journals.aom.org/doi/10.5465/amd.2021.0144.summary | 3 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,678 | rp888 | 2024-11-02T11:23:38 | null | null | null | 1 | null | [
42025679
] | null | true | null | null | null | null | null | null | null | train |
42,025,680 | Turboblack | 2024-11-02T11:23:51 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,710 | luizsantana | 2024-11-02T11:30:50 | Show HN: I automated Next.js deploy to Hetnzer VPS | Got a bit tired of unpredictable Vercel bills and created a way to automatically deploy in a Hetzner VPS.<p>- Commit push deploys to VPS<p>- Easy setup of a VPS<p>- SSL and Nginx config automatically<p>Made it easy to:<p>1. login with github<p>2. add hetzner api key<p>3. create VPS<p>4. select github app<p>5. add DNS<p>6. voila, app deployed. Every new commit updates with blue/green deploy<p>also handy to add redis or whatever else to same VPS | https://productive.dev | 4 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,714 | katkova7584 | 2024-11-02T11:31:42 | null | null | null | 1 | null | [
42025715
] | null | true | null | null | null | null | null | null | null | train |
42,025,723 | zeristor | 2024-11-02T11:34:30 | Tears for Fears – Dancing to Their Own Trauma ("Mad World") (Trash Theory) [video] | null | https://www.youtube.com/watch?v=Ke3xktAJgyg | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,724 | skacekamen | 2024-11-02T11:34:32 | null | null | null | 1 | null | [
42025725
] | null | true | null | null | null | null | null | null | null | train |
42,025,732 | rntn | 2024-11-02T11:36:08 | Financial institutions told to clean house before next CrowdStrike strikes | null | https://www.theregister.com/2024/11/02/fca_it_resilience/ | 3 | 0 | null | null | null | no_error | Financial institutions told to get their house in order before the next CrowdStrike strikes | 2024-11-02T09:30:08Z | Connor Jones |
The UK's finance regulator is urging all institutions under its remit to better prepare for IT meltdowns like that of CrowdStrike in July.
The Financial Conduct Authority (FCA) said issues at unregulated third parties were the leading cause of operational disruption within Blighty's financial institutions between 2022 and 2023.
Many major organizations were affected to varying degrees by CrowdStrike's software cockup over the summer, including some of the world's leading banks and trading houses.
JPMorgan Chase's trade execution systems were reportedly affected, some Bloomberg terminals were rendered inaccessible, the London Stock Exchange was hit, and ION Group, UBS, CMC Markets, and others also all reported issues.
"These outages emphasize firms' increasing dependence on unregulated third parties to deliver important business services," the FCA said in a statement. "This highlights the importance of firms continuing to become operationally resilient in line with our rules.
"We encourage all firms, regardless of how they were affected by the CrowdStrike incident, to consider these lessons, to improve their ability to respond to and recover from future disruptions."
For those of you who somehow missed out on what will be remembered as one of the defining IT events of 2024, back in July, CrowdStrike pushed a now-infamous channel file update to its Falcon EDR platform. That update contained a critical logic error, causing Falcon to crash so hard that Windows did too, displaying blue screens of death on 8.5 million PCs worldwide. A bad time was had by many trying to fix this.
Soon, many financial institutions in the UK will be forced by the FCA to become resilient to these kinds of events. The regulator's rules (PS21/3) governing third-party events like CrowdStrike's, requiring in-scope organizations to implement robust business continuity measures that mitigate the worst impacts of incidents like IT outages, came into force in March 2022. The deadline to become compliant – March 2025 – is fast approaching.
The FCA said those who had already met the requirements of PS21/3 demonstrated the best response to the CrowdStrike outage. They were able to effectively prioritize which systems to bring back online first, minimizing the operational impact on the business and wider market, as well as consult prepared incident response and communications plans.
If they mapped their systems and third-party relationships, organizations demonstrated a stronger ability to manage their exposure to limit the overall impact of the incident.
From a technical perspective, some affected institutions were forced to identify single points of failure in their tech stacks and make changes accordingly. For example, some sought alternative products or operating systems, while others decided to review their change management processes relating to software updates.
The FCA urged all regulated organizations to ensure their update-testing procedures were up to scratch and amend them where necessary so any faults can be contained more easily. This especially applies to institutions whose services are relied upon by other key players in the industry.
Delta officially launches lawyers at $500M CrowdStrike problem
CrowdStrike's Blue Screen blunder: Could eBPF have saved the day?
CrowdStrike apologizes to Congress for 'perfect storm' that caused global IT outage
1 in 10 orgs dumping their security vendors after CrowdStrike outage
Other recommendations included preparing external comms templates, such as website banners so all customers and stakeholders are comprehensively informed about any issues in a timely manner. Plus, the usual incident response preparations you'd typically expect any organization to have in place.
Despite the widespread impact on financial markets, the institutions involved largely got on with things and recovered relatively quickly. Little fuss has been made of the incident since.
The same can't be said for Delta Air Lines, however, which recently launched legal proceedings against CrowdStrike, looking to recoup at least some of the circa $500 million in revenue it claims to have lost thanks to the outage.
Delta faced significant challenges, taking longer than most to return to service. It blamed CrowdStrike and Microsoft, and in response they pointed the finger straight back, saying the airline refused their offers of free technical support.
CrowdStrike also alleged Delta was running on aging IT equipment, a major factor in why it took so long to recover.
Shortly after Delta filed its lawsuit against the cybersecurity company, CrowdStrike itself launched a counter-suit alleging "Delta's own negligence" led to the issues it faced. ®
| 2024-11-08T12:08:54 | en | train |
42,025,744 | orbesargentina | 2024-11-02T11:39:23 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,745 | latexr | 2024-11-02T11:39:40 | iPod fans evade Apple's DRM to preserve 54 lost clickwheel-era games | null | https://arstechnica.com/gaming/2024/11/ipod-fans-evade-apples-drm-to-preserve-54-lost-clickwheel-era-games/ | 5 | 1 | [
42025755,
42026110,
42025840
] | null | null | null | null | null | null | null | null | null | train |
42,025,750 | jbdaily | 2024-11-02T11:40:53 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,764 | geox | 2024-11-02T11:44:15 | Neurotechnology boosts memory without surgery | null | https://actu.epfl.ch/news/neurotechnology-boosts-memory-without-surgery/ | 9 | 0 | [
42026098
] | null | null | null | null | null | null | null | null | null | train |
42,025,765 | axijiajiajia | 2024-11-02T11:44:25 | null | null | null | 4 | null | [
42025766
] | null | true | null | null | null | null | null | null | null | train |
42,025,782 | rustoo | 2024-11-02T11:47:43 | Solid-state batteries enter pilot production, costs expected to drastically drop | null | https://www.pv-magazine.com/2024/11/01/solid-state-batteries-enter-pilot-production-costs-expected-to-drastically-drop/ | 30 | 4 | [
42026014,
42032761,
42026094
] | null | null | null | null | null | null | null | null | null | train |
42,025,785 | nature556 | 2024-11-02T11:48:06 | Fact [pdf] | null | https://ranjitjhala.github.io/static/fact_dsl.pdf | 9 | 4 | [
42025786,
42066133,
42064707
] | null | null | null | null | null | null | null | null | null | train |
42,025,806 | bookofjoe | 2024-11-02T11:50:54 | Anti-Aging Enthusiasts Are Taking a Pill to Extend Their Lives. Will It Work? | null | https://www.nytimes.com/2024/09/24/well/live/rapamycin-aging-longevity-benefits-risks.html | 4 | 1 | [
42025808
] | null | null | null | null | null | null | null | null | null | train |
42,025,812 | francescoxx | 2024-11-02T11:52:03 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,025,813 | Tomte | 2024-11-02T11:52:19 | 21,000 workers dead in 8 years of Mohammed bin Salman's 'Saudi vision 2030' | null | https://www.hindustantimes.com/world-news/21000-workers-dead-in-8-years-of-mohammed-bin-salmans-saudi-vision-2030-report-101730127065962.html | 18 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,025,815 | stevejm | 2024-11-02T11:52:43 | Faster matrix multiplication through reinforcement learning | null | https://deepnote.com/workspace/katkas-workspace-5e77d5a7-62ef-41b0-8a6b-5521efccf6c0/project/Discovering-faster-matrix-multiplication-algorithms-with-reinforcement-learning-12af6cdc-429a-4706-84c2-0b4cebe56deb/notebook/5f0022194a564e4fa66db6713ae74fe3 | 9 | 1 | [
42025816,
42025829
] | null | null | null | null | null | null | null | null | null | train |
42,025,817 | PatrickPushes | 2024-11-02T11:52:53 | I built a no-code map app builder for iOS and Android | null | https://blitzwizmap.com | 4 | 0 | [
42025818
] | null | null | null | null | null | null | null | null | null | train |
42,025,838 | sagasu007 | 2024-11-02T11:58:03 | null | null | null | 1 | null | [
42026083,
42025961
] | null | true | null | null | null | null | null | null | null | train |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.