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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
42,012,321 | geekmetaverse | 2024-10-31T22:44:35 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,012,347 | michaefe | 2024-10-31T22:51:39 | Mapping Application for Penguin Populations and Projected Dynamics | null | https://www.penguinmap.com/mapppd/ | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,354 | dmckinno | 2024-10-31T22:52:47 | The Dunning-Kruger Effect in LLMs | null | https://www.ddmckinnon.com/2024/10/31/dans-weekly-ai-speech-and-language-scoop-32/ | 2 | 0 | null | null | null | missing_parsing | Dan's Weekly AI Speech and Language Scoop #32 - Daniel D. McKinnon | 2024-10-31T22:51:47+00:00 | On SimpleQA and factuality in LLMsOpenAI released an LLM benchmark called SimpleQA [csv of questions]. It is far from simple. This is pub trivia on steroids. I suspect I could reason my way to directionally correct answers for about 10% of them and don’t even fully understand what the question is referring to at least 25% of the time.The best performing model, o1-preview, only answers 42.7% of the questions correctly while the worst, Claude-3-haiku, only scores 5.1%. Bad, right? Not so fast. While Claude-3-haiku only answers 5.1% correctly, it also only answers 19.6% incorrectly, passing on 75.3%. o1-preview misses 48.1%, making its correct answer rate worse than a coin flip.Which would you prefer as a user? I personally would rather Haiku. When I am interacting with an LLM, especially over audio, I want the answer to be correct. I don’t want the additional mental overhead of having to check the 50% of answers. I want the model to know that the answers to these questions are probably not lurking in its parametric memory and to seek out the answer for me.And existing LLM-based products for the most part do just this. From sampling a half dozen questions Meta AI, ChatGPT, and Gemini all score nearly perfectly on this benchmark with their search grounding (tragically, Claude does not offer this yet; surprised that they are releasing agents but still haven’t built search into their consumer product).So why is this an LLM rather than user-facing system eval? Why do I care about how the base model scores on these very challenging trivia questions? The answer, I think, is that I don’t. This is very reminiscent of the early work on teaching LLMs to do complex arithmetic. People used to care until they realized the correct approach was to just give the model a calculator.However, OpenAI motivates this benchmark in an interesting way. They suggest that it can be used to measure the calibration (confidence) of language models, which is quite interesting and requires a set of prompts that the model is likely to struggle with.I’m always glad to see an interesting new benchmark, but I hope that this doesn’t start popping up in marketing materials as something to optimize for and instead remains an interesting way to explore model calibration. LLMs can do lots of wonderful things and I don’t believe that recalling obscure facts is one of their capabilities to optimize for. I would much rather see a smaller, faster, more agentic model than one that has memorized Wikipedia.Stated vs. revealed LLM preferences: Dunning-Kruger Effect in LLMsFigure 2 in the SimpleQA paper shows how the OpenAI researchers measured model calibration. They both asked the model how confident it was in the answer and sampled the sample prompt multiple times to measure how frequently it yielded the same answer. This is very reminiscent of a human stated vs. revealed preference experiment.When asked (stated preference), the models were alway far more confident in their answers than justified (only answered 50% of the questions correctly for which they were 90% confident). However, when sampled, the models were much better calibrated (and o1-preview almost perfectly so). They are just like us humans: deludedly overconfident in our abilities. The information theory behind parametric memoryHow many parameters do you need to memorize Wikipedia? According to some very smart people who write papers I don’t totally understand on the physics of language models, the empirical limit for knowledge storage in LLMs is 2 bits per parameter. English Wikipedia contains 28.2B characters. While optimal Huffman code for the English alphabet requires 4 bits per character, Wikipedia tells me that this actually drops to ~1 bit/character due to the redundancies of the language.This means that we need 14.1B parameters to just memorize Wikipedia. And Wikipedia is a small fraction of the information found on the internet and only had answers to a fraction (~50%?) of the questions in SimpleQA. This means that we would need ~28.2B parameters to just store the answers to these questions at the theoretical limit. I suspect the real number is far higher.Reasoning one’s way to winning pub triviaHave you ever found yourself sitting at a bar with friends playing trivia and totally stumped by the answer? What was Lift 5 at Beaver Creek called in 2005? What is lift 5 today? Is that Rose Bowl? Cinch? Who knows? Let’s have a drink.If you’re anything like me, reasoning through true trivia questions rarely works. You either know the number of rushing yards some obscure NFL player had in a given year or you don’t (no points for Fermi estimates in trivia!).OpenAI’s previously mentioned SimpleQA eval shows that machines are no different. Their GPT-40 model scores 38% correct and o1-preview only adds 5 pp through reasoning. This is in sharp contrast to true reasoning problems like competition math where o1-preview adds a whopping 44 pp.This is another nice demonstration that models are just like us afterall. | 2024-11-08T21:20:13 | null | train |
|
42,012,355 | YuriRDev | 2024-10-31T22:53:15 | Scan Any Website for Vulnerabilities in One Click For Free | null | https://www.defendia.app/ | 3 | 0 | [
42012356
] | null | null | null | null | null | null | null | null | null | train |
42,012,387 | pjebs | 2024-10-31T22:57:28 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,012,397 | chagaif | 2024-10-31T22:59:06 | Flutter Has Been Forked | null | https://medium.com/@sachindas246/flutter-has-been-forked-256faea67f7c | 1 | 1 | [
42012455
] | null | null | null | null | null | null | null | null | null | train |
42,012,400 | LinuxBender | 2024-10-31T22:59:31 | Cloud repatriation officially a trend for specific workloads | null | https://www.theregister.com/2024/10/30/cloud_repatriation_about_specific_workloads/ | 6 | 0 | [
42012680
] | null | null | null | null | null | null | null | null | null | train |
42,012,421 | LinuxBender | 2024-10-31T23:02:06 | Beijing claims it's found underwater lighthouses that its foes use for espionage | null | https://www.theregister.com/2024/10/30/china_seabed_surveillance_device_claims/ | 4 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,435 | flyaway123 | 2024-10-31T23:05:13 | Something Happens When You Take This to Its Logical Conclusion | null | https://www.youtube.com/watch?v=Q10_srZ-pbs | 3 | 1 | [
42012732
] | null | null | no_article | null | null | null | null | 2024-11-07T23:21:11 | null | train |
42,012,441 | irqlevel | 2024-10-31T23:06:13 | Show HN: Grammar checker for text on website pages | null | https://sitechecker.cc/ | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,461 | LinuxBender | 2024-10-31T23:09:08 | Honeypot: Researchers Catch Attackers Exposing 15K Stolen Credentials in S3 | null | https://www.securityweek.com/honeypot-surprise-researchers-catch-attackers-exposing-15000-stolen-credentials-in-s3-bucket/ | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,467 | AlchemistCamp | 2024-10-31T23:09:36 | Russia fines Google twenty decillion dollars | null | https://www.cnn.com/2024/10/31/tech/google-fines-russia/index.html | 3 | 1 | [
42012495
] | null | null | null | null | null | null | null | null | null | train |
42,012,509 | cdvonstinkpot | 2024-10-31T23:15:20 | ML-based bee recognition and tracking for advancing insect behavior research | null | https://link.springer.com/article/10.1007/s10462-024-10879-z | 2 | 0 | null | null | null | body_too_long | null | null | null | null | 2024-11-08T16:01:19 | null | train |
42,012,514 | mgl | 2024-10-31T23:15:34 | Ask HN: Seeking connections with insurance and insurtech companies | Dear Friends,
we are seeking connections with insurance and insurtech companies interested in building and delivering faster and we’d love to gather their feedback on our open-source platform:<p>https://openkoda.com/<p>I would be grateful for your hints and introductions, thank you! | null | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,527 | petethomas | 2024-10-31T23:18:01 | AI Models Replace Real People in Mango's Fast-Fashion Ads | null | https://www.bloomberg.com/news/articles/2024-10-31/mango-clothing-uses-ai-to-replace-some-fashion-models-in-ads | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,530 | kristopolous | 2024-10-31T23:18:36 | Russia fines Google $2.5 decillion for YouTube bans | null | https://fortune.com/europe/2024/10/31/russia-fines-google-2-5-decillion-dollars-youtube-bans/ | 3 | 1 | [
42012599
] | null | null | null | null | null | null | null | null | null | train |
42,012,539 | sheriffofpaddys | 2024-10-31T23:20:42 | A Glass Slipper is a Bad Way to Find Cinderella | null | https://divingintheshallowend.com/if-the-shoe-fits/ | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,564 | geox | 2024-10-31T23:24:35 | Instagram-famous squirrel seized by New York state authorities | null | https://apnews.com/article/new-york-pet-squirrel-seized-peanut-pnut-c2baa345d9e76416e0ca6d6e947ee272 | 5 | 0 | null | null | null | no_error | Instagram-famous squirrel named Peanut seized by New York state authorities | 2024-10-31T19:16:13 | By
KAREN MATTHEWS |
A New York man who turned a rescued squirrel into a social media star called Peanut is pleading with state authorities to return his beloved pet after they seized it during a raid that also yielded a raccoon named Fred.Multiple anonymous complaints about Peanut — also spelled P’Nut or PNUT — brought at least six officers from the state Department of Environmental Conservation to Mark Longo’s home near the Pennsylvania border in rural Pine City on Wednesday, Longo said.“The DEC came to my house and raided my house without a search warrant to find a squirrel!” said Longo, who is 34. “I was treated as if I was a drug dealer and they were going for drugs and guns.”The officers left with Peanut, who amassed hundreds of thousands of followers on Instagram, TikTok and other platforms during his seven years with Longo. They also took Fred, a more recent addition to the family.
A spokesperson for the DEC said in a statement that the agency started an investigation after receiving “multiple reports from the public about the potentially unsafe housing of wildlife that could carry rabies and the illegal keeping of wildlife as pets.”
Longo, who runs an animal refuge inspired by his squirrel buddy called P’Nuts Freedom Farm Animal Sanctuary, took to Instagram to mourn Peanut’s loss.“Well internet, you WON,” Longo posted. “You took one of the most amazing animals away from me because of your selfishness. To the group of people who called DEC, there’s a special place in hell for you.”
Longo fears that Peanut has been euthanized. “I don’t know if Peanut is alive,” he said in a phone interview Thursday. “I don’t know where he is.”The DEC spokesperson did not respond to a question about whether Peanut had been euthanized.Longo said he saw Peanut’s mother get hit by a car in New York City seven years ago, leaving the tiny squirrel an orphan. Longo brought Peanut home and cared for him for eight months before trying to release the squirrel into the great outdoors. “A day and a half later I found him sitting on my porch missing half of his tail with his bone sticking out,” Longo said.
Longo determined that Peanut lacked the survival skills to live in the wild and would remain an indoor squirrel. Soon after Longo posted videos of Peanut playing with his cat, internet fame followed.A scroll through Peanut’s Instagram account suggests that this is no ordinary squirrel. Peanut leaps on to Longo’s shoulder, he wears a miniature cowboy hat, he eats a waffle while wearing crocheted bunny ears.Over the years Peanut’s story has been featured on TV and newspapers including USA Today.Longo, who works as a mechanical engineer, was living in Norwalk, Connecticut, until he decided to move to upstate New York last year to start an animal sanctuary. P’Nuts Freedom Farm Animal Sanctuary opened in April 2023 and now houses about 300 animals including horses, goats and alpacas, said Longo, who runs the sanctuary with his wife, Daniela, and other family members.Longo is aware that it’s against New York state law to own a wild animal without a license. He said he was in the process of filing paperwork to get Peanut certified as an educational animal.
“If we’re not following the rules, guide us in the right direction to follow the rules, you know?” Longo said. “Let us know what we need to do to have Peanut in the house and not have to worry about him getting taken.”As for Fred, Longo said he only had the raccoon for a few months and was hoping to rehabilitate the injured creature and release him back to the woods.Longo is not the first animal owner to protest the confiscation of a pet by New York authorities. A Buffalo-area man whose alligator was seized by the DEC in March is suing the agency to get the 750-pound (340-kilogram) reptile back.
| 2024-11-07T23:58:38 | en | train |
42,012,586 | wtznc | 2024-10-31T23:27:33 | The Raspberry Pi Security Guide [pdf] | null | http://www.nhatqbui.com/assets/TheUltimateRaspberryPiSecurityGuide.pdf | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,603 | pjebs | 2024-10-31T23:30:05 | null | null | null | 1 | null | [
42012604
] | null | true | null | null | null | null | null | null | null | train |
42,012,615 | pjebs | 2024-10-31T23:32:33 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,012,623 | PaulHoule | 2024-10-31T23:33:40 | Merlin Solar bets twisty panels will help it land on rooftops everywhere | null | https://techcrunch.com/2024/10/27/merlin-solar-bets-twisty-panels-will-help-it-land-on-rooftops-everywhere/ | 4 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,632 | mrcgnc | 2024-10-31T23:35:05 | Memecoins Are the Point | null | http://jpkoning.blogspot.com/2024/10/memecoins-are-point.html | 3 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,659 | teleforce | 2024-10-31T23:40:40 | Python for DevOps Lecture and Workshop | null | https://github.com/noahgift/python-for-devops-april-2022 | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,799 | eduardoleao052 | 2024-11-01T00:06:06 | PyTorch in the Browser (JavaScript) | null | https://github.com/eduardoleao052/js-pytorch | 1 | 1 | [
42012800
] | null | null | null | null | null | null | null | null | null | train |
42,012,807 | uxhacker | 2024-11-01T00:07:04 | Researchers are questioning if ADHD should be seen as a disorder | null | https://www.economist.com/science-and-technology/2024/10/30/researchers-are-questioning-if-adhd-should-be-seen-as-a-disorder | 29 | 39 | [
42013723,
42013290,
42019490,
42012990,
42014857,
42013473,
42018620,
42017941,
42013314,
42013305,
42013385,
42018929,
42018592,
42013295,
42018582,
42028382,
42013412,
42013260
] | null | null | null | null | null | null | null | null | null | train |
42,012,823 | GCA10 | 2024-11-01T00:10:12 | The Rise and Fall of the German Stock Market, 1870-1938 [pdf] | null | https://www.econsoc.hist.cam.ac.uk/docs/CWPESH_number_25_Sept_2016.pdf | 1 | 0 | null | null | null | is_pdf | null | null | null | null | 2024-11-08T11:18:33 | null | train |
42,012,834 | pedrosbmartins | 2024-11-01T00:12:01 | KSA, the Kerbal Space Program Replacement from RocketWerkz | null | https://old.reddit.com/r/KerbalSpaceProgram/comments/1gg5106/ksa_the_ksp_replacement_from_rocketwerkz_seamless/ | 4 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,835 | joebig | 2024-11-01T00:12:27 | Robur the Conqueror | null | https://en.wikipedia.org/wiki/Robur_the_Conqueror | 2 | 1 | [
42012843
] | null | null | null | null | null | null | null | null | null | train |
42,012,856 | petethomas | 2024-11-01T00:16:23 | The economy is going great, except for the housing market | null | https://www.investopedia.com/the-economy-is-going-great-except-for-one-huge-problem-8737911 | 32 | 83 | [
42013079,
42013107,
42013411,
42013160,
42013186,
42013167,
42017283,
42013207
] | null | null | no_error | The Economy Is Going Great, Except For One Huge Problem | null | Full Bio | Key Takeaways
High prices and mortgage rates have crushed the housing market, even as other parts of the economy have prospered.A long-standing housing shortage has contributed to the housing crunch, alongside the Federal Reserve's anti-inflation rate hikes, which drove up mortgage rates.Mortgage rates could ease next year, improving the outlook but not necessarily solving the underlying problems.
Amid a recent spate of data showing the economy humming along smoothly, one sector has consistently stayed out of whack and it’s a big one: housing.On Thursday, a report from the Bureau of Economic Analysis showed rising housing costs have helped core inflation run higher than the Federal Reserve’s target 2% annual rate. A drop-off in homebuilding in the third quarter was a drag on the economy’s overall growth rate, according to a report from the bureau on the Gross Domestic Product Wednesday. Earlier in the month, data showed homebuilding languished and sales of existing houses skidded to their lowest in more than a decade in September.The beleaguered housing market is a stark contrast to other important pillars of the economy, which are running relatively smoothly: the job market is holding on to an extended winning streak, consumers are spending freely, and inflation is falling.
At the heart of the problem is the fact that high prices and high mortgage rates have pushed the cost of buying a house out of reach for people with typical incomes who previously could afford the payments. A mortgage payment on the typical newly bought home would take up 42% of a household's median income in August, up from 29% in January 2020, according to a housing affordability monitor created by the Federal Reserve Bank of Atlanta.
Housing problems are wreaking havoc on the overall economy, hurting family budgets and preventing people from moving to take advantage of job opportunities, among other ripple effects.
High Mortgage Rates Hurt Budgets
The housing market has been collateral damage in the Federal Reserve’s war on inflation.“When the Fed raises interest rates, it hits the housing industry hardest because It's the most interest rate sensitive sector,” Bill Adams, chief economist at Comerica Bank, said in an interview with Investopedia.Mortgage rates hit record lows during the pandemic, as the Fed held its influential benchmark interest rate near zero to boost the economy. But when the Fed raised rates rapidly in 2022 to combat inflation, mortgage rates surged. By October 2023, the average rate offered for a 30-year mortgage hit a two-decade high of 7.79%, up from the record low of 2.65% in January 2021, according to Freddie Mac.
The whiplash virtually paralyzed the housing market, as homeowners who secured ultra-low rates during the pandemic hesitated to sell their homes and exchange them for new mortgages at higher rates.As a result of that “lock-in” effect, there are far fewer homes on the market than before the pandemic, according to the National Association of Realtors. And homebuilders haven’t been keeping up with the demand for new homes, partly because local zoning regulations restrict the construction of new houses where they’re most in demand. And on top of that, demand for bigger homes surged during the pandemic as workers adjusted to the new telecommuting lifestyle.“You've had this huge step change in the amount of living space that Americans want since the pandemic because so many more Americans work from home now or work from home on some days,” Adams said. “Demand for living space in the United States is just permanently higher than it was pre-pandemic. It’s the flip side of all those empty offices in big city downtowns.”All those forces have combined to keep prices repeatedly hitting record highs even though many buyers have been priced out of the market.
What's Next?
One part of that equation—mortgage rates—could improve in the near future.
The Federal Reserve cut the fed funds rate from a two-decade high in September and plans further rate cuts in the coming months as inflation simmers down to its target of a 2% annual rate. While mortgage rates don’t always move in tandem with the fed funds rate, they are influenced by it. Forecasters at Fannie Mae expect mortgage rates to drop to the mid-5% range by the end of next year, compared to 6.72% as of last week.“2024 was a really tough year for the housing market,” Adams said. “2025 should be a better year because the Fed is cutting.”However, that outlook could be shifted depending on which party wins Tuesday's general election. Both presidential candidates have touted plans to ease the housing shortage, with Vice President Kamala Harris promising to build 3 million affordable homes and former president Donald Trump pledging to free up homes by deporting immigrants.
| 2024-11-07T08:13:31 | en | train |
42,012,889 | todsacerdoti | 2024-11-01T00:24:05 | Control Characters in ASCII and Unicode | null | https://www.aivosto.com/articles/control-characters.html | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,912 | delichon | 2024-11-01T00:27:53 | Strange Bedfellows: Native American Tribes and Big Pharma (2019) | null | https://scholarship.law.duke.edu/dlj/vol68/iss7/3/ | 4 | 1 | [
42012946
] | null | null | null | null | null | null | null | null | null | train |
42,012,913 | jpkoning | 2024-11-01T00:28:03 | Econ 101 is wrong about tariffs | null | https://www.economicforces.xyz/p/econ-101-is-wrong-about-tariffs | 7 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,917 | todsacerdoti | 2024-11-01T00:28:22 | Checking Linearizability in Go | null | https://notes.eatonphil.com/2024-10-31-checking-linearizability-in-go.html | 3 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,928 | joeyjiron06 | 2024-11-01T00:29:38 | Spooky Launch: Unveiling a New Tool for Web Developers This Halloween | Hey HN community,<p>Get ready for a treat (no tricks, I promise)! This Halloween, I’m thrilled to launch my new tool for web developers. My AI-powered platform helps you conjure up beautiful, responsive landing pages in seconds using Tailwind CSS. Say goodbye to the horror of endless coding and design!
I’m in the beta phase and need your magical insights to make this tool even better. As a Halloween special, I’m offering limited free access to the beta. Don’t ghost me – join the beta and help me brew up the perfect tool for developers!
Creep on over to landmarkai.dev and sign up for free.<p>Thanks for your support, and happy coding!<p>Joey Jiron | null | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,938 | cynicalpeace | 2024-11-01T00:30:59 | AI Generated Game: Oasis | null | https://twitter.com/Etched/status/1852089772329869436 | 14 | 3 | [
42013237,
42015569
] | null | null | null | null | null | null | null | null | null | train |
42,012,974 | smitec | 2024-11-01T00:39:07 | Lessons from one too many model spreadsheets | null | https://www.elliotcsmith.com/lessons-from-one-too-many-model-spreadsheets/ | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,012,999 | md_rumpf | 2024-11-01T00:44:17 | Writing Code Just-in-Time | null | https://maxrumpf.com/writing/2024-08-21-jit-coding.html | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,000 | doublepg23 | 2024-11-01T00:44:46 | The Mainstreaming of Loserdom | null | https://tellthebeees.substack.com/p/the-mainstreaming-of-loserdom | 5 | 0 | null | null | null | no_error | The Mainstreaming of Loserdom | 2024-08-15T16:41:06+00:00 | Tell the Bees | I know the title is provocative, but let me make my case first. Over the past few years, something has shifted in the perception of acceptable recreational behavior, or the way people talk about their hobbies: people are gleeful to admit they have no hobbies, no interests, no verve. Somehow, one of the main “hobbies” accepted by the masses is staying home, laying in bed, scrolling on their phones and watching television.What happened?Not to channel my queen Lexi Featherstone from “Splat!” (One of the most memorable episodes of Sex and the City ) but it does seem like no one’s fun anymore online. Earlier this month, Charli XCX had a birthday party as part of Brat Summer and thousands of chronically online Twitter users rushed to proclaim that they never partied as a weird form of moral superiority.75,000 people agreed with this 😔A TikTok of a woman saying her favorite place to be was in bed got 28 million views, with thousands of people agreeing that all they cared about was being home/being in bed. Another TikToker stated that his favorite place in the world to be was in bed watching television, and he didn’t need to go to the club at all.My issue isn’t with the action, per se (although when soothsayer Glamdemon2004 said “name three hobbies you have outside of media consumption” three years ago people went absolutely nuts and bullied her off TikTok) it’s with the perception of it. In the nineties, going out was the height of glamour. The SATC girls were always partying. The characters on Buffy and Charmed were always at the club. The culture yearned for Brat Summer because we’ve been fossilized, idolizing being at home. It was clear twenty years ago that someone who rarely engaged with their peers, didn’t really have friends, and didn’t really leave their house wasn’t aspirational: they were odd. Shot, chaser.I know what people are going to say: not everyone drinks, not everyone parties, we have social anxiety, everything is too expensive (a future Substack is about our current Economic Crisis and the performance of poverty online). I don’t want to blame the phones, but I will say: the dating apps are in crisis. People aren’t meeting in person. There’s a singles wall in New York City, famously one of the easiest cities to be single in. People are taking up running to meet one another. We’re in a loneliness crisis. People simply aren’t connecting the way they used to, and I won’t be the bad guy for pointing out that it doesn’t surprise me that people are desperately lonely while also saying their favorite hobby is… staying home.I’ll also defend myself preemptively and say not everyone has the same threshold for social interaction, which again, is fine. My issue is that I do not believe that the millions of people engaging with these posts all have very literal tolerance for social interaction. There is the issue of an individualistic culture and declining social etiquette (posts about people stating “I hate my coworkers and I hate small talk and I refuse to engage in any conversation outside my job” routinely go viral, and I think that’s a failing of etiquette, sorry!) allowing people to state these opinions without shame, but it does feel like a larger societal failing that the activity people crave the most is scrolling on their phones, watching other people live their lives.Not all is lost: some of Gen Z realize something is going on.I’ve been on the internet for twenty years: I’ve been on fanfiction.net, I’ve been on Livejournal, I’ve been on Tumblr. I was surrounded by people who spent time alone, but they were creating. They were writing, they were generating, they were knitting and sewing and painting and dreaming. The specific activity I’m talking about is a lack of any of this. The people screaming from their rooftops about how they don’t go anywhere and don’t have any friends aren’t the same people writing 70,000 words of Harry/Draco smut, I’m sorry! I know my people, and this feels different. It feels more sinister. Posting fanfiction online is a bid for community. Scrolling on your phone is not.The discourse around Brat Summer was fascinating. Someone tweeted that Charli probably didn’t party the way she claimed to in her music, that she had a 15 step skincare routine and went to bed by 9. It was a joke, but it was also a failure of imagination: people couldn’t believe that she was actually partying. People couldn’t believe she was outside for real. When a video surfaced of her doing drugs at a rave, Gen Z melted down. How could she? (I will point out that the most repeated lyric in “365” is bumping that, another lyrics asks should we do a little key? Should we have a little line?) Shot, chaser v2.Another viral post asked if Brat Summer was just about watching rich people have fun, with some saying “they’re not actually having fun, they wish they were on their phones”. I promise you, some people are actually having fun! There is fun to be found outside of the phone!There are some signs of life from the younger generation. I recently saw a post asking why people were shaming others who did enjoy going to the club, while this post asked if she was the only 24 year old who didn’t enjoy going to be at 9 PM and dressing like someone’s aunt. It feels like a psyop: the people who are out partying aren’t online, so the online folks have coordinated this mass push to make us believe partying is bad and staying home alone is the solution.On social media, we’ve seen this belief that once you get a partner or turn 25 the “adult” thing to do is to be home with them every night, and that going out is childish. On Twitter a few weeks ago, someone was saying that this is an odd belief: there are entire jobs that rely on people networking, going out, and “partying”, well into higher levels of responsibility. I’ve been to my fair share of literary events, I’ve worked at advertising agencies, I’ve seen people much older than me getting down. It’s not looked down upon because it’s part of the job, and the people hired for those roles are much more inclined to crave that level of social interaction. There isn’t some switch that flips when you turn 25 that makes you believe leaving your house is ungodly. Putting all of this together, I do think a combination of loneliness, social anxiety, unlimited internet access and economic crisis are possibly combining into something. We have viral TikToks of saying “people think I’m depressed because I’m always in bed” and a viral tweet of people clamoring for an oddly shaped couch bed, promising they’d never leave their house if they could own it (“96,000 of you are Jack Russell terriers” made me laugh out loud). I don’t think it’s out of the realm of possibility. If someone is saying the only thing they want to do is sleep, it’s not strange to wonder if something else might be going on. As Buffy once said, “I didn’t jump to conclusions. I simply took a step and there conclusions were.”I promise I’m not attacking: I’m questioning. I do think an over-identification with the introvert label (as both Millennials and Gen Z are wont to do) and the rise of bed rotting and the ease of opting out of society has created an illusory community around… not doing anything. I don’t think these people need to be shamed (I’m sorry for the title of this post, but it’s too good to change) but I wonder if they need to be brought into community. People aren’t going to church, they aren’t joining civic groups, they think posting online is activism, they aren’t making friends, or their friends live too far away, or their friends are married, etc. When I wrote about situationships, I came to the same conclusion: loneliness makes us do strange things. Whether staying with someone who treats your heart like garbage or creating an online campaign to make us believe being home is the answer, I think a lot of posting is a genuine cry for the warmth of other people. If you need me, I’ll be outside. | 2024-11-08T02:01:40 | en | train |
42,013,005 | mavelikara | 2024-11-01T00:45:01 | The Defense Reformation | null | https://18theses.com/ | 3 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,017 | zaven | 2024-11-01T00:48:10 | Typing "boo " into ChatGPT makes a little ghost pop up | null | https://chatgpt.com/ | 3 | 0 | [
42013018
] | null | null | missing_parsing | ChatGPT | null | null | | 2024-11-08T05:10:47 | null | train |
42,013,025 | zeristor | 2024-11-01T00:49:09 | Capturing carbon from the air just got easier | null | https://news.berkeley.edu/2024/10/23/capturing-carbon-from-the-air-just-got-easier/ | 5 | 0 | null | null | null | no_error | Capturing carbon from the air just got easier - Berkeley News | 2024-10-23T15:00:00+00:00 | Robert Sanders |
A new type of porous material called a covalent organic framework quickly sucks up carbon dioxide from ambient air
The new porous material for capturing carbon dioxide, called a covalent organic framework (COF), has hexagonal channels decorated with polyamines that efficiently bind carbon dioxide molecules (blue and orange balls) at concentrations found in ambient air.
Chaoyang Zhao
October 23, 2024
Capturing and storing the carbon dioxide humans produce is key to lowering atmospheric greenhouse gases and slowing global warming, but today’s carbon capture technologies work well only for concentrated sources of carbon, such as power plant exhaust. The same methods cannot efficiently capture carbon dioxide from ambient air, where concentrations are hundreds of times lower than in flue gases.
Yet direct air capture, or DAC, is being counted on to reverse the rise of CO2 levels, which have reached 426 parts per million (ppm), 50% higher than levels before the Industrial Revolution. Without it, according to the Intergovernmental Panel on Climate Change, we won’t reach humanity’s goal of limiting warming to 1.5 °C (2.7 °F) above preexisting global averages.
A new type of absorbing material developed by chemists at the University of California, Berkeley, could help get the world to negative emissions. The porous material — a covalent organic framework (COF) — captures CO2 from ambient air without degradation by water or other contaminants, one of the limitations of existing DAC technologies.
“We took a powder of this material, put it in a tube, and we passed Berkeley air — just outdoor air — into the material to see how it would perform, and it was beautiful. It cleaned the air entirely of CO2. Everything,” said Omar Yaghi, the James and Neeltje Tretter Professor of Chemistry at UC Berkeley and senior author of a paper that will appear online Oct. 23 in the journal Nature.
“I am excited about it because there’s nothing like it out there in terms of performance. It breaks new ground in our efforts to address the climate problem,” he added.
A vial of COF-999, which is yellow, with UC Berkeley’s landmark campanile in the background.Zihui Zhou, UC Berkeley
According to Yaghi, the new material could be substituted easily into carbon capture systems already deployed or being piloted to remove CO2 from refinery emissions and capture atmospheric CO2 for storage underground.
UC Berkeley graduate student Zihui Zhou, the paper’s first author, said that a mere 200 grams of the material, a bit less than half a pound, can take up as much CO2 in a year — 20 kilograms (44 pounds) — as a tree.
“Flue gas capture is a way to slow down climate change because you are trying not to release CO2 to the air. Direct air capture is a method to take us back to like it was 100 or more years ago,” Zhou said. “Currently, the CO2 concentration in the atmosphere is more than 420 ppm, but that will increase to maybe 500 or 550 before we fully develop and employ flue gas capture. So if we want to decrease the concentration and go back to maybe 400 or 300 ppm, we have to use direct air capture.”
COF vs MOF
Yaghi is the inventor of COFs and MOFs (metal-organic frameworks), both of which are rigid crystalline structures with regularly spaced internal pores that provide a large surface area for gases to stick or adsorb. Some MOFs that he and his lab have developed can adsorb water from the air, even in arid conditions, and when heated, release the water for drinking. He has been working on MOFs to capture carbon since the 1990s, long before DAC was on most people’s radar screens, he said.
UC Berkeley graduate student Zihui Zhou with a 100 milligram test sample of COF-999. The sample was placed in the analyzer behind Zhou to measure carbon dioxide adsorption from an air mixture similar to that of ambient air.Robert Sanders, UC Berkeley
Two years ago, his lab created a very promising material, MOF-808, that absorbs CO2, but the researchers found that after hundreds of cycles of adsorption and desorption, the MOFs broke down. These MOFs were decorated inside with amines (NH2 groups), which efficiently bind CO2 and are a common component of carbon capture materials. In fact, the dominant carbon capture method involves bubbling exhaust gases through liquid amines that capture the carbon dioxide. Yaghi noted, however, that the energy intensive regeneration and volatility of liquid amines hinders their further industrialization.
Working with colleagues, Yaghi discovered why some MOFs degrade for DAC applications — they are unstable under basic, as opposed to acidic, conditions, and amines are bases. He and Zhou worked with colleagues in Germany and Chicago to design a stronger material, which they call COF-999. Whereas MOFs are held together by metal atoms, COFs are held together by covalent carbon-carbon and carbon-nitrogen double bonds, among the strongest chemical bonds in nature.
As with MOF-808, the pores of COF-999 are decorated inside with amines, allowing uptake of more CO2 molecules.
“Trapping CO2 from air is a very challenging problem,” Yaghi said. “It’s energetically demanding, you need a material that has high carbon dioxide capacity, that’s highly selective, that’s water stable, oxidatively stable, recyclable. It needs to have a low regeneration temperature and needs to be scalable. It’s a tall order for a material. And in general, what has been deployed as of today are amine solutions, which are energy intensive because they’re based on having amines in water, and water requires a lot of energy to heat up, or solid materials that ultimately degrade with time.”
Yaghi and his team have spent the last 20 years developing COFs that have a strong enough backbone to withstand contaminants, ranging from acids and bases to water, sulfur and nitrogen, that degrade other porous solid materials. The COF-999 is assembled from a backbone of olefin polymers with an amine group attached. Once the porous material has formed, it is flushed with more amines that attach to NH2 and form short amine polymers inside the pores. Each amine can capture about one CO2 molecule.
Omar Yaghi with molecular models of some of his porous structures, called metal-organic frameworks, or MOFs. COFs have similar internal structures, but are held together by strong covalent bonds instead of by metal atoms.Brittany Hosea-Small for UC Berkele
When 400 ppm CO2 air is pumped through the COF at room temperature (25 °C) and 50% humidity, it reaches half capacity in about 18 minutes and is filled in about two hours. However, this depends on the sample form and could be speeded up to a fraction a minute when optimized. Heating to a relatively low temperature — 60 °C, or 140 °F — releases the CO2, and the COF is ready to adsorb CO2 again. It can hold up to 2 millimoles of CO2 per gram, standing out from other solid sorbents.
Yaghi noted that not all the amines in the internal polyamine chains currently capture CO2, so it may be possible to enlarge the pores to bind more than twice as much.
“This COF has a strong chemically and thermally stable backbone, it requires less energy, and we have shown it can withstand 100 cycles with no loss of capacity. No other material has been shown to perform like that,” Yaghi said. “It’s basically the best material out there for direct air capture.”
Yaghi is optimistic that artificial intelligence can help speed up the design of even better COFs and MOFs for carbon capture or other purposes, specifically by identifying the chemical conditions required to synthesize their crystalline structures. He is scientific director of a research center at UC Berkeley, the Bakar Institute of Digital Materials for the Planet (BIDMaP), which employs AI to develop cost-efficient, easily deployable versions of MOFs and COFs to help limit and address the impacts of climate change.
“We’re very, very excited about blending AI with the chemistry that we’ve been doing,” he said.
The work was funded by King Abdulaziz City for Science and Technology in Saudi Arabia, Yaghi’s carbon capture startup, Atoco Inc., Fifth Generation’s Love, Tito’s, and BIDMaP. Yaghi’s collaborators include Joachim Sauer, a visiting scholar from Humboldt University in Berlin, Germany, and computational scientist Laura Gagliardi from the University of Chicago.
RELATED INFORMATION
Carbon dioxide capture from open air using covalent organic frameworks (Nature)
Omar Yaghi’s website
| 2024-11-08T02:32:49 | en | train |
42,013,032 | neitsab | 2024-11-01T00:50:42 | VMware Workstation Shifting from Proprietary Code to Using Upstream KVM | null | https://www.phoronix.com/news/VMware-Workstation-KVM | 50 | 40 | [
42013708,
42013162,
42013246,
42015043,
42013459,
42013596,
42013390,
42013610,
42013930,
42013122,
42014406
] | null | null | no_error | VMware Workstation Shifting From Proprietary Code To Using Upstream KVM | null | Written by Michael Larabel in Virtualization on 31 October 2024 at 08:57 PM EDT. 45 Comments |
This isn't an off-schedule April Fools' Joke or anything like that but an exciting sign of the times: VMware Workstation will be shifting off its proprietary base and onto leveraging the upstream Linux Kernel-based Virtual Machine (KVM) for virtualization needs moving forward.
This week a kernel patch series hit the Linux kernel mailing list by Broadcom (formerly VMware) engineer Zack Rusin entitled KVM: x86: Small changes to support VMware guests. That patch series explicitly lays out:
"To be able to switch VMware products running on Linux to KVM some minor changes are required to let KVM run/resume unmodified VMware guests.
First allow enabling of the VMware backdoor via an api. Currently the setting of the VMware backdoor is limited to kernel boot parameters, which forces all VM's running on a host to either run with or without the VMware backdoor. Add a simple cap to allow enabling of the VMware backdoor on a per VM basis. The default for that setting remains the kvm.enable_vmware_backdoor boot parameter (which is false by default) and can be changed on a per-vm basis via the KVM_CAP_X86_VMWARE_BACKDOOR cap.
Second add a cap to forward hypercalls to userspace. I know that in general that's frowned upon but VMwre guests send quite a few hypercalls from userspace and it would be both impractical and largelly impossible to handle all in the kernel. The change is trivial and I'd be maintaining this code so I hope it's not a big deal."
Beyond the patch cover letter explicitly mentioning "to be able to switch VMware products running on Linux to KVM", I have been able to confirm directly with Broadcom that this indeed is in motion. VMware Workstation for desktop virtualization will indeed be shifting from their existing proprietary virtualization code and begin making use of the widely-used KVM code.
However, Broadcom doesn't yet have a timeline for making this transition. In part that transition is going to depend upon how quickly their necessary KVM changes can be upstreamed and in turn picked up by the major Linux distributions. So with that said it will likely be well into 2025 at a minimum before an official transition given the length of the kernel merge windows and even longer before the non-rolling-release distributions upgrade to new kernel versions. But this move for VMware Workstation switching to KVM is indeed happening.
I'm not sure I ever envisioned VMware abandoning their proprietary virtualization code in favor of leveraging upstream KVM but in any event it's a terrific success story for the upstream KVM community. VMware Workstation will presumably remain a commercial/proprietary product (but already with free personal use) as opposed to transitioning like a full open-source project but in any event to see their key proprietary virtualization code move in favor of the open-source KVM code is a huge accomplishment.Update: Broadcom would like to clarify that while using KVM for the CPU virtualization, they will continue to rely on all of the existing VMware virtual devices for graphics and other functionality. Also on both macOS and Windows they have migrated to the native CPU virtualization frameworks. | 2024-11-08T00:19:29 | en | train |
42,013,044 | 2OEH8eoCRo0 | 2024-11-01T00:52:53 | Elon Musk Changed Course to Go All Out for Trump | null | https://www.nytimes.com/2024/10/31/business/elon-musk-trump.html | 8 | 3 | [
42015426,
42013084,
42013116
] | null | null | null | null | null | null | null | null | null | train |
42,013,053 | orientalseed | 2024-11-01T00:54:47 | Dongzi Securities Big Data Research Report on Public Opinion | null | https://dongzisecurity.com/index.html | 1 | 1 | [
42013054
] | null | null | null | null | null | null | null | null | null | train |
42,013,062 | PaulHoule | 2024-11-01T00:56:11 | Vegetable growers in Australia deploy 'good bugs' to reduce pesticide use | null | https://www.abc.net.au/news/2024-10-10/good-bugs-reduce-vegetable-pesticides-glasshouse-virginia-sa/104371086 | 10 | 1 | [
42013863
] | null | null | null | null | null | null | null | null | null | train |
42,013,066 | hassaniiiis | 2024-11-01T00:57:47 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,013,071 | xqcgrek2 | 2024-11-01T00:58:14 | Brunner's stick mantis has no males | null | https://en.wikipedia.org/wiki/Brunneria_borealis | 3 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,086 | cebert | 2024-11-01T01:02:48 | If you invested $10k any of these stocks in 1999, you'd have over $1M now | null | https://www.cnbc.com/2024/10/25/nvidia-apple-what-a-10k-investment-25-years-ago-would-be-worth-now.html | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,091 | bookofjoe | 2024-11-01T01:04:01 | A conserved fertilization complex bridges sperm and egg in vertebrates | null | https://www.cell.com/cell/fulltext/S0092-8674(24)01093-6?_returnURL=https%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS0092867424010936%3Fshowall%3Dtrue | 37 | 18 | [
42016787,
42016595,
42014827,
42013684,
42014801,
42016967,
42013950
] | null | null | null | null | null | null | null | null | null | train |
42,013,095 | williswee | 2024-11-01T01:04:39 | BYD surpasses Tesla in Q3 earnings | null | https://www.techinasia.com/news/byd-surpasses-tesla-q3-earnings | 4 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,109 | surprisetalk | 2024-11-01T01:07:56 | Automat | null | https://automat.org | 4 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,117 | IrisBMeredith | 2024-11-01T01:09:33 | null | null | null | 10 | null | [
42013525,
42013414,
42013474,
42013488
] | null | true | null | null | null | null | null | null | null | train |
42,013,123 | runnably | 2024-11-01T01:10:43 | Autonomous Starlink-powered R/C boat [video] | null | https://www.youtube.com/watch?v=lmj0qeOVcsk | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,127 | sandwichsphinx | 2024-11-01T01:11:17 | Canada and Mexico Fear a More Protectionist America | null | https://www.wsj.com/economy/trade/us-election-free-tade-mexico-canada-95ff98dc | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,136 | transpute | 2024-11-01T01:12:16 | Linux Boot Time Improvement Presentations (2006-2024) | null | https://elinux.org/Boot_Time_Presentations | 2 | 0 | null | null | null | no_error | Boot Time Presentations - eLinux.org | null | null | Contents
1 Accelerating Linux Kernel Boot-Up for Large Multi-Core Systems [LPC 2024]
2 Productizing the Linux boot time tweaks and tricks – an engineering problem ! [LPC 2024]
3 Boot time testing with ftrace [LPC 2024]
4 Initiatives in boot time reduction - boot time markers, boot phases and profile-guided optimizations [LPC 2024]
5 Accelerating Linux Boot Time: Techniques and Strategies for Optimal Performance [DevConf.US 2024]
6 Deferred Memblocks Init for Boot Time Reduction [ELC 2024]
7 Software Optimization for Early Linux Boot for Time Critical Applications [ELC 2024]
8 Unlocking the Potential of Suspend to RAM Using Linux in a Multi-Core, Multi-Firmware Automotive SoC [ELC 2024]
9 Resolve and standardize early access to hardware for automotive industry with Linux [LPC 2023]
10 Improving kexec boot time [LPC 2023]
11 The Resurrection of Ureadahead and Speeding up the Boot Process and Preloading Applications [ELCE 2023]
12 Boot-Time Optimization for the Real World [ELCE 2020]
13 Timing Boot Time Reduction Techniques [ELCE 2019]
14 We Need to Talk About Systemd: Boot Time Optimization for the New init daemon [ELCE 2019]
15 Introduction to NAND Flash Aware Hibernation-based Boot [ELC 2019]
16 Improving Embedded Systems Boot Time by Hibernation: An Overview on the State of the Art and a Case of Study on i.MX family of Processors [ELC 2019]
17 Secure Updates for a Memory Constrained XIP System [ELC 2019]
18 Introduction to Memory Constrained XIP System [ELC 2019]
19 BoF LinuxBoot: Linux as Firmware [ELC 2018]
20 A Pragmatic Guide to Boot-Time Optimization [ELCE 2017]
21 Improving the Bootup Speed of AOSP [ELC 2017]
22 How to Boot Linux in One Second [ELC 2015 ELCE 2015]
23 Fastboot Tools and Techniques [ELC 2015]
24 Tutorial: Useful systemd Functionalities Without systemd [ELCE 2015]
25 Update on Boot Time Reduction Techniques with Figures [ELC 2014]
26 Linux Quickboot [ELC 2014]
27 12 Lessons Learnt in Boot Time Reduction [ELCE 2014]
28 Boot Time Optimizations [ELCE 2012]
29 Snapshot Booting on Embedded Linux [ELC 2011]
30 Faster Resume For More Energy Savings on MeeGo [ELC 2011]
31 Linux without a boot loader [ELC 2010]
32 The Right Approach to Minimal Boot Times [ELCE 2010]
33 Barebox: Booting Linux Fast and Fancy [ELCE 2010]
34 Lightweight Prelinker for Kernel Modules [ELCE 2010]
35 Quantitative analysis of system initialization in embedded Linux systems [ELC 2009]
36 Update on Boot Time Reduction Techniques [ELCE 2009]
37 How We Got a 3D Application Booting in 5 Seconds Under Linux [ELCE 2009]
38 LKM Fast Loader Based on ELF Hash Table [ELCE 2009]
39 Tools and Techniques for Reducing Bootup Time [ELCE 2008]
40 Bootup Technologies BOF [ELC 2007]
41 Parallelizing Linux boot on CE Devices [ELCE 2007]
42 Linux Suspend-to-Disk Objectives for Consumer Electronic Devices [ELCE 2007]
43 Visualizing resource usage during initialization of embedded systems [ELC 2006]
44 Boot-Time Optimization - results of applying currently available solutions [ELC 2006]
Accelerating Linux Kernel Boot-Up for Large Multi-Core Systems [LPC 2024]
Presenters: Saurabh Singh Sengar, Srivatsa Bhat, Microsoft
Summary:
This talks described optimizations to boot machines with thousands of CPUs. Techniques used are
deferring calculations, and parallelizing CPU hotplug to bring up cores asynchronously instead of
sequentially.
Session page | PDF | VIDEO
Productizing the Linux boot time tweaks and tricks – an engineering problem ! [LPC 2024]
Presenter: Khasim Syed Mohammed, Texas Instruments
Summary:
This talk presents some of the challenges to productizing boot time optimizations, including
handling combined use cases, accellerated MCU cores and "Linux late attach" (handing pre-initialized hardware
off to the kernel), and determining how to package documentation and fixes for optimizations for
delivery to customers.
Session page | | PDF | VIDEO
Boot time testing with ftrace [LPC 2024]
Presenter: Laura Nao, Collabora
Summary:
This session discusses how to leverage existing tracing tools to
monitor key events in the boot sequence and proposes building a kselftest to
automate this process. This involves identifying critical functions to trace,
configuring ftrace (e.g. through bootconfig), parsing trace data, and
comparing it to user-provided reference values to detect significant boot
time regressions.
Session page | PDF | VIDEO
Initiatives in boot time reduction - boot time markers, boot phases and profile-guided optimizations [LPC 2024]
Presenter: Tim Bird, Sony
Summary:
This session discusses 1) boot time markers, 2) boot phases, and 3) profile-guided boot-time optimizations.
Boot-time markers is a proposed set of well-define measurement points in the Linux boot process, used for testing improvements and regressions in boot time. "Boot phases" refers to dividing the kernel boot process into two distinct phases: a time-critical phase and non-time-critical (or deferred) phase, Profile-guided boot-time optimizations refers to utilizing run-time data from one instantiation of the kernel to drive the optimization of subsequent instantiations.
Session page | | PDF | VIDEO
Accelerating Linux Boot Time: Techniques and Strategies for Optimal Performance [DevConf.US 2024]
Presenters: Eric Curtin, Ed Chong, Brian Masney RedHat
Summary:
The presentation provided a variety of strategies and techniques
to optimize boot time, from measuring boot performance to specific
optimizations within systemd, kernel, and filesystem configurations.
The presentation discusses tools such as systemd-analyze, and
several specific optimizations including expedited RCU, use of
EROFS, and minimizing initramfs.
Session page | PDF | VIDEO
Deferred Memblocks Init for Boot Time Reduction [ELC 2024]
Presenter: Sudarshan Rajagopalan, Qualcomm
Summary:
This talk presents work to reduce boot time by bringing up the system with
just a subset of memory (say 1G). Then, after kernel/userspace init,
the remaining blocks of RAM are initialized in an asynchronous parallel fashion
using the memory hotplug framework in the Linux kernel (utilizing the parallelism of the SMP system).
This is done using kthreads within the kernel or by a userspace service.
On systems with large RAM, this saves up to 200 ms (20 to 30 ms per GB).
PDF | VIDEO
Software Optimization for Early Linux Boot for Time Critical Applications [ELC 2024]
Presenters: Aashvij Shenai, Soumya Tripathy, Texas Instruments
Summary:
This talk covers requirements in automotive for quick kernel boot and I/O: audio
chime in 200ms, a splash screen in 500ms, a Linux kernel boot in 1s and a display
app in 1.5s.
Optimizations in the areas of PMIC choice, bootmedia, ROM, bootloaders, OS kernel,
filesystem and user-space applications are discussed. Different methods of initializing hardware
and handing off control between heterogeneous cores is presented.
The stripped down Linux kernel is tuned carefully to avoid removing any
configs required for a fast boot. Several kernel drivers have their probe order changed
to avoid deferrals. A tiny filesystem is used. The boot time was reduced from over 25 seconds
to about 1 second.
PDF | VIDEO
Unlocking the Potential of Suspend to RAM Using Linux in a Multi-Core, Multi-Firmware Automotive SoC [ELC 2024]
Presenter: Gregory Clement, Bootlin
Summary:
This talk discusses Suspend to RAM (S2R) on a sophisticated SoC featuring diverse CPU cores, and multiple
firmware components, including open-source ones like TF-A, U-Boot, and OP-TEE. This implementation handles
a case where the SoC is entirely powered off during S2R. TF-A memory in SRAM is stored to DDR, which is put
in self-refresh mode. On restore, TF-A and DM firmware images are reloaded.
PDF | VIDEO
Resolve and standardize early access to hardware for automotive industry with Linux [LPC 2023]
Presenter: Khasim Syed Mohammed, Texas Instruments
Summary:
The presentation covers automotive use cases where audio, graphics, CAN bus, ethernet
and camera streams need to come up very quickly (between 100ms and 500ms from cold boot).
Current solutions are presented, and there is a proposal to standardize methods for
"Linux late attach", where hardware is initialized by firmware and handed off to
the Linux kernel.
Session page | PDF | VIDEO
Improving kexec boot time [LPC 2023]
Presenter: Usama Arif, TikTok
Summary:
This session covers work done to reduce kernel boot times in a typical server configuration
by up to 80%, including parallelizing smpboot, optimizing TSC synchronization, and the work
in progress to streamline the initialization of struct pages when using HugeTLB Vmemmap
Optimization (HVO). Additionally, optimizations such as skipping PCI probe for certain
devices and skipping purgatory are covered. Other areas which still take significant boot
time, such as enabling ACPI interpreter and ACPI table loads are also discussed.
Session page | PDF | VIDEO
The Resurrection of Ureadahead and Speeding up the Boot Process and Preloading Applications [ELCE 2023]
Presenter: Steven Rostedt, Google
Summary:
The history and current status of ureadahead is discussed. Google has been maintaining a version as part of the ChromeOS project. Steve explains what ureadahead used to do as well as what he has done to improve it. ureadhead can be used to parallelize pre-fetching of pages for an executable image. In ChromeOS testing, ureadeahead saved about 1.2 seconds (about 14.5% reduction) of boot time.
Session Page | PDF | VIDEO
Boot-Time Optimization for the Real World [ELCE 2020]
Presenter: Michael Olbrich, Pengutronix e.K.
Summary:
There have been many talks about boot-time optimization in the past. For the most part, the only goal was a minimum boot-time. It's a good way to show the optimization techniques and gives a nice wow effect. But the end result is rarely usable in real world projects. So instead of looking for new ways to reduce the boot-time just a little bit more, this talk will look at boot-time optimization in a larger context. "The device needs to boot faster" is often stated but the actual requirements behind this are often more complex. We will look at typical requirements and possible solutions beyond general boot-time optimization. And while booting as fast as possible is nice, in most cases there are other more important requirements, which are often related to security or reliability. This talk will look at possible optimizations in this context. What are the consequences of an optimization and what trade-offs are possible? And last but not least, let's talk about the hardware. The choices made while designing a device can have a big impact on the boot-time. Therefore this talk will end with advices for hardware design criteria and component decisions to keep in mind to boot fast.
Session page | PDF | VIDEO
Timing Boot Time Reduction Techniques [ELCE 2019]
Presenter: Michael Opdenacker, Bootlin
Summary:
The talk presented methods to optimize kernel boot time, such as toolchain options,
kernel compression options, kernel command line parameters, eliminating unnecessary
copying in the bootloader, optimizing storage performance and using U-Boot's Falcon mode.
Michael also showed tools to identify parts which can be eliminated (such as finding all
the files not accessed during the boot sequence) or optimized. Total techniques reduced
boot time (to first camera frame displayed) from 9.45 seconds to 2.4 seconds.
Session page | PDF | VIDEO
We Need to Talk About Systemd: Boot Time Optimization for the New init daemon [ELCE 2019]
Presenter: Chris Simmonds, 2net
Summary:
This talk gives and introduction to systemd, and systemd-analyze, and shows how to eliminate services that are not used for a critical app.
Chris compares the speed and complexity of busybox, System V init, and systemd init processes. Initial trimming reduced boot time by 35 seconds.
PDF | VIDEO
Introduction to NAND Flash Aware Hibernation-based Boot [ELC 2019]
Presenter: Kyungsik Lee, LG Electronics
Summary:
The slides cover existing boot time reduction and
hibernation boot techniques and talks about new proposal
for hibernation boot.
PDF | VIDEO
Improving Embedded Systems Boot Time by Hibernation: An Overview on the State of the Art and a Case of Study on i.MX family of Processors [ELC 2019]
Presenter: Nicola La Gloria & Laura Nao, Kynetics LLC
Summary:
This Presentation will provide the current state of the
kernel development of Hibernation on ARM architectures
and details tests on the popular i.MX family of processors.
PDF | VIDEO
Secure Updates for a Memory Constrained XIP System [ELC 2019]
Presenter: Vitaly Wool, Konsulko
Summary:
This session details updates using XIP stands for eXecute In Place
– a technology that allows code to be executed directly from flash
without copying the code to RAM first.
PDF | VIDEO
Introduction to Memory Constrained XIP System [ELC 2019]
Presenter: Vitaly Wool, Konsulko
Summary:
This presentation provides information about the execution of code directly
from flash without copying the code to ram first.
PDF | VIDEO
BoF LinuxBoot: Linux as Firmware [ELC 2018]
Presenter: Chris Koch & Gan Shun, Google
Summary:
This presentation covers methods to reduce drivers and firmware in a Linux system for efficient and faster boot.
PDF
A Pragmatic Guide to Boot-Time Optimization [ELCE 2017]
Presenter: Chris Simmonds
Summary:
This presentation covers
From power-on to starting qtdemo app boot was reduced from 11.76 seconds to 2.83 seconds, using a variety of techniques.
Tools shown were: grabserial, bootchart, bootgraph. Reduction techniques shown were: running the app as init, removing unused
filesystem (BTRFS), adding 'quiet' to command line, removing unused drivers, reducing zImage size, removing u-boot bootdelay,
and using u-boot falcon mode.
PDF | VIDEO
Improving the Bootup Speed of AOSP [ELC 2017]
Presenter: Bernhard Rosenkranzer, Linaro
Summary:
This presentation provides information about the
approaches to improve time from cold start to UI.
PDF | VIDEO
How to Boot Linux in One Second [ELC 2015 ELCE 2015]
Presenter: Jan Altenberg, Linutronix, GmbH
Summary:
This presentation covers some technical basic, optimizations
in boot-loader, kernel, file-system and application.
ARMv5 based device is taken as a use-case.
PDF
Fastboot Tools and Techniques [ELC 2015]
Presenter: John Mehaffey, Mentor Graphics
Summary:
This presentation talks fastboot, a mechanism
to boot the Embedded system faster.
PDF
Tutorial: Useful systemd Functionalities Without systemd [ELCE 2015]
Presenter: Bartosz Golaszewski, BayLibre
Summary:
This presentation talks about ACME, sigrok and faster boot.
PDF | VIDEO
Update on Boot Time Reduction Techniques with Figures [ELC 2014]
Presenter: Michael Opdenacker, Free Electrons
Summary:
This session covers extensive details about Linux
boot time reduction methodology and resources.
PDF | VIDEO
Linux Quickboot [ELC 2014]
Presenter: Tristan Lelong, Adeneo Embedded
Summary:
This session covers different mechanisms that exist
to optimize boot time and ways to integrate them to
achieve the best time on a custom board with fully
loaded kernel running.
PDF | VIDEO
12 Lessons Learnt in Boot Time Reduction [ELCE 2014]
Presenter: Andrew Murray, Embedded Bits Limited
Summary:
This presentation explains different ways to reduce boot time.
Also describes pros and cons of those methods.
PDF
Boot Time Optimizations [ELCE 2012]
Presenter: Alexandre Belloni, Adeneo Embedded
Summary:
This presentation talks about boot time optimizations,
it's use cases, kernel compression, SMP and file systems.
PDF
Snapshot Booting on Embedded Linux [ELC 2011]
Presenter: Kang Dongwook (ETRI)
Summary:
This presentation discusses about a fastboot mechanism
called snapshot booting for achieving faster boot times.
PDF
Faster Resume For More Energy Savings on MeeGo [ELC 2011]
Presenter: Yoshiya Hirase (Nokia)
Summary:
This presentation discusses about hibernation to
handset devices for more energy saving.
PDF
Linux without a boot loader [ELC 2010]
Presenter: Greg Ungerer
Summary:
This session explores options to boot Linux faster
without the boot loader.
PDF | ODP | SOURCE
The Right Approach to Minimal Boot Times [ELCE 2010]
Presenter: Andrew Murray
Summary:
This presentation gives information about principles
behind boot time reduction, an approach to boot time reduction
with a Case study on " MS7724 'Ecovec'".
PDF
Barebox: Booting Linux Fast and Fancy [ELCE 2010]
Presenter: Robert Schwebel and Sascha Hauer
Summary:
This presentation provides information about
Barebox project, it's history, development resources,
hardware support and also Linux booting sequence.
PDF
Lightweight Prelinker for Kernel Modules [ELCE 2010]
Presenter: Carmelo Amoroso and Rosario Contarino
Summary:
This presentation gives information about fast boot, kernel
modules and Prelinker. It states pros and cons of Prelinker
and it's strategy.
PDF
Quantitative analysis of system initialization in embedded Linux systems [ELC 2009]
Presenter: Andre Puschmann
Summary:
This presentation analyzes time taken by different layers
during boot time and initialization.
PDF
Update on Boot Time Reduction Techniques [ELCE 2009]
Presenter: Michael Opdenacker
Summary:
This presentation discusses various techniques to reduce
boot time such as LZO kernel decompression, jffs2, disabling
console output, disabling auto IP config etc.
PDF
How We Got a 3D Application Booting in 5 Seconds Under Linux [ELCE 2009]
Presenter: Grégory Clement
Summary:
This presentation talks about the boot process, time measurement
techniques for optimizations.
PDF
LKM Fast Loader Based on ELF Hash Table [ELCE 2009]
Presenter: Carmelo Amoroso
Summary:
This presentation discusses various strategies to speed-up
kernel boot times with a modular kernel.
PDF
Tools and Techniques for Reducing Bootup Time [ELCE 2008]
Presenter: Tim Bird
Summary:
This presentation discusses about the tools used such as
grab-serial for system-wide measurement, bootchart, strace
and process trace that can be used for boot time reduction.
PDF | VIDEO
Bootup Technologies BOF [ELC 2007]
Presenter: Elias Kesh
Summary:
These slides discuss about different charters of
Boot Technology Work Group (BTWG).
PDF
Parallelizing Linux boot on CE Devices [ELCE 2007]
Presenter: Vitaly Wool
Summary:
These slides give information on boot up time, boot time
profiling, sequence driver init, asynchronous driver init.
PDF | VIDEO
Linux Suspend-to-Disk Objectives for Consumer Electronic Devices [ELCE 2007]
Presenter: Vitaly Wool
Summary:
This presentation discusses about Suspend-to-Disk(STD),
compares with Suspend-to-Ram(STR), pros and cons of STD,
and also use case for CE devices.
PDF | VIDEO
Visualizing resource usage during initialization of embedded systems [ELC 2006]
Presenter: Matthew Klahn
Summary:
This presentation does a quantitative analysis of resource
utilization during booting and initialization.
PDF
Boot-Time Optimization - results of applying currently available solutions [ELC 2006]
Presenter: Prof. Nicholas Mc Guire
Summary:
The presentation discusses various optimizations that
can be applied in the kernel for faster boot time.
PDF
| 2024-11-08T15:51:09 | en | train |
42,013,137 | cloudking | 2024-11-01T01:12:37 | Ask HN: Any tools to detect AI generated music? | Is anyone aware of tools that can help detect AI generated music? | null | 2 | 1 | [
42013224
] | null | null | null | null | null | null | null | null | null | train |
42,013,140 | geekkong | 2024-11-01T01:13:02 | Halloween 2048 | null | https://2048.love/2048Halloween | 3 | 0 | [
42013141
] | null | null | null | null | null | null | null | null | null | train |
42,013,150 | mariusandra | 2024-11-01T01:15:11 | FrameOS NimConf 2024 | null | https://frameos.net/blog/nimconf-2024/ | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,165 | flybird | 2024-11-01T01:18:45 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,013,205 | begoon | 2024-11-01T01:26:54 | PEP 750 – Template Strings | null | https://peps.python.org/pep-0750/ | 1 | 1 | [
42019531
] | null | null | null | null | null | null | null | null | null | train |
42,013,219 | T-Winsnes | 2024-11-01T01:29:31 | Does Alpine resolve DNS properly? (2021) | null | https://purplecarrot.co.uk/post/2021-09-04-does_alpine-resolve_dns_properly/ | 2 | 1 | [
42016181
] | null | null | null | null | null | null | null | null | null | train |
42,013,225 | davikr | 2024-11-01T01:30:06 | I got a new machine. Windows 11. It's stuck updating. It is taking ages. Why? | null | https://twitter.com/s4schoener/status/1852022939807342910 | 4 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,231 | signa11 | 2024-11-01T01:30:47 | Demystifying the Regular Expression That Checks If a Number Is Prime (2016) | null | https://illya.sh/the-codeumentary-blog/regular-expression-check-if-number-is-prime/ | 1 | 1 | [
42013243
] | null | null | null | null | null | null | null | null | null | train |
42,013,250 | Karen10 | 2024-11-01T01:33:26 | null | null | null | 1 | null | [
42013251
] | null | true | null | null | null | null | null | null | null | train |
42,013,258 | hoerzu | 2024-11-01T01:35:14 | What's missing from this guide "ChatGPT explained for people who don't use it"? | null | https://franz101.substack.com/p/chatgpt-explained-for-people-who | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,296 | luellafletcher0 | 2024-11-01T01:42:04 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,013,298 | veggieWHITES | 2024-11-01T01:42:13 | Let's Encrypt was unable to validate domain names | null | https://letsencrypt.status.io/pages/55957a99e800baa4470002da | 19 | 10 | [
42013422,
42013692,
42013654,
42013645
] | null | null | null | null | null | null | null | null | null | train |
42,013,313 | gmays | 2024-11-01T01:45:46 | First Black Hole in a Triple Star System Found | null | https://www.caltech.edu/about/news/first-black-hole-in-a-triple-star-system-found | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,328 | HeliumHydride | 2024-11-01T01:48:17 | Watch out for unexpected performance increases as well as decreases | null | https://devblogs.microsoft.com/oldnewthing/20121123-00/?p=6013 | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,336 | eugenhotaj | 2024-11-01T01:51:01 | Spelled out implementation of LLM parallelisms in pure C | null | https://github.com/EugenHotaj/ml.c/blob/main/parallelisms/README.md | 3 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,367 | kamaraju | 2024-11-01T01:57:29 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,013,368 | sandwichsphinx | 2024-11-01T01:57:40 | DAWN: Designing Distributed Agents in a Worldwide Network | null | https://arxiv.org/abs/2410.22339 | 10 | 3 | [
42013993,
42014348,
42014163
] | null | null | null | null | null | null | null | null | null | train |
42,013,371 | mcharawi | 2024-11-01T01:58:19 | Electric planemaker Beta Technologies raises over $300M in new funding | null | https://www.reuters.com/business/aerospace-defense/electric-planemaker-beta-technologies-raises-over-300-mln-new-funding-2024-10-31/ | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,379 | afdbcreid | 2024-11-01T01:59:50 | Feds: Critical Software Must Drop C/C++ by 2026 or Face Risk | null | https://thenewstack.io/feds-critical-software-must-drop-c-c-by-2026-or-face-risk/ | 45 | 89 | [
42013923,
42013558,
42013788,
42013888,
42015121,
42013575,
42013415,
42013749,
42013871,
42013908,
42013714,
42013816,
42014042,
42014195,
42013865,
42013731,
42013436,
42013828,
42013528
] | null | null | null | null | null | null | null | null | null | train |
42,013,419 | michellezzz | 2024-11-01T02:06:00 | Deploy a PostgreSQL Cluster on Kubernetes with high availability | null | https://kubeblocks.io/docs/preview/user_docs/kubeblocks-for-postgresql/cluster-management/create-and-connect-a-postgresql-cluster | 2 | 0 | null | null | null | no_error | Create and connect to a PostgreSQL Cluster | KubeBlocks | null | null | This tutorial shows how to create and connect to a PostgreSQL cluster.Create a PostgreSQL clusterBefore you startInstall kbcli if you want to manage the PostgreSQL cluster by kbcli.Install KubeBlocks by kbcli or by Helm.Make sure the PostgreSQL Addon is enabled. The PostgreSQL Addon is installed and enabled by KubeBlocks by default. But if you disable it when installing KubeBlocks, enable it first.kbclikubectlkbcli addon list>NAME TYPE STATUS EXTRAS AUTO-INSTALL ...postgresql Helm Enabled true...View all the database types and versions available for creating a cluster.kbclikubectlkbcli clusterdefinition listkbcli clusterversion listTo keep things isolated, create a separate namespace called demo throughout this tutorial.kubectl create namespace demoCreate a clusterKubeBlocks supports creating two types of PostgreSQL clusters: Standalone and Replication Cluster. Standalone only supports one replica and can be used in scenarios with lower requirements for availability. For scenarios with high availability requirements, it is recommended to create a Replication Cluster, which creates a cluster with a Replication Cluster to support automatic failover. To ensure high availability, Primary and Secondary are distributed on different nodes by default.kbclikubectlCreate a PostgreSQL cluster.Here is an example of creating a Standalone.kbcli cluster create postgresql mycluster -n demokbcli provides various options for you to customize your cluster specifications, such as setting cluster version, termination policy, CPU, and memory. You can view these options by adding --help or -h flag.kbcli cluster create postgresql --helpkbcli cluster create postgresql -hFor example, you can create a Replication Cluster with the --replicas flag.kbcli cluster create postgresql mycluster --replicas=2 -n demoIf you only have one node for deploying a Replication Cluster, set the --topology-keys as null when creating a Replication Cluster. But you should note that for a production environment, it is not recommended to deploy all replicas on one node, which may decrease the cluster availability.kbcli cluster create postgresql mycluster --replicas=2 --availability-policy='none' -n demoVerify whether this cluster is created successfully.kbcli cluster list -n demo>NAME NAMESPACE CLUSTER-DEFINITION VERSION TERMINATION-POLICY STATUS CREATED-TIMEmycluster demo postgresql postgresql-14.8.0 Delete Running Sep 28,2024 16:47 UTC+0800Connect to a PostgreSQL Clusterkbclikubectlport-forwardkbcli cluster connect mycluster --namespace demoFor the detailed database connection guide, refer to Connect database. | 2024-11-07T07:58:22 | en | train |
42,013,427 | Judyrabbit | 2024-11-01T02:07:06 | Low Code Written in the Grid | null | https://blog.scudata.com/low-code-written-in-the-grid/ | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,428 | lewis106355 | 2024-11-01T02:07:16 | null | null | null | 1 | null | [
42013429
] | null | true | null | null | null | null | null | null | null | train |
42,013,430 | 8jef | 2024-11-01T02:07:34 | Cosmic Desktop Alpha 3 Released with More Enhancements | null | https://www.phoronix.com/news/COSMIC-Desktop-Alpha-3 | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,433 | lisper | 2024-11-01T02:08:50 | The Final Experiment – Antarctica | null | https://www.the-final-experiment.com/ | 2 | 0 | null | null | null | missing_parsing | The Final Experiment - Antarctica | null | null | Close xThank you! Your submission has been received!Oops! Something went wrong while submitting the form.
| 2024-11-08T21:01:21 | null | train |
42,013,441 | lando2319 | 2024-11-01T02:11:44 | null | null | null | 7 | null | [
42013618,
42013791,
42022685
] | null | true | null | null | null | null | null | null | null | train |
42,013,445 | Kokouane | 2024-11-01T02:13:01 | Apex Legends removes their game from Linux due to cheating concerns | null | https://twitter.com/PlayApex/status/1852019667315102151 | 3 | 1 | [
42013466
] | null | null | null | null | null | null | null | null | null | train |
42,013,453 | PaulHoule | 2024-11-01T02:14:42 | A simplified form of life: Scientists build modules for a synthetic cell | null | https://phys.org/news/2024-10-life-scientists-modules-synthetic-cell.html | 5 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,475 | mooreds | 2024-11-01T02:18:35 | Getting Started with Claude Computer Use | null | https://riza.io/blog/claude-computer-use | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,478 | impish9208 | 2024-11-01T02:19:14 | Parents of 2 year-old Bitten By Rattlesnake Charged Nearly $300k | null | https://www.msn.com/en-us/health/medical/pure-greed-parents-of-2yo-bitten-by-rattlesnake-charged-nearly-300k-for-life-saving-cure/ar-AA1thJOx | 5 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,495 | thunderbong | 2024-11-01T02:22:54 | The Most Similar Countries in the World | null | https://objectivelists.com/the-objectively-most-similar-countries/ | 1 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,511 | hcavarsan | 2024-11-01T02:26:07 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,013,524 | bilater | 2024-11-01T02:28:42 | Show HN: My Simple Lofi Player | As AI takes over our jobs, we can focus on the important stuff like building our own Lofi players.<p>I added some neat features like sound effects - rain, cafe ambience, people talking, and keyboard typing - that you can play with and customize your lofi experience.<p>Always wanted to have my own little lofi player and I'm going to have some great coding sessions with this hopefully.<p>Appreciate any feedback! | https://www.you-tldr.com/lofi | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,530 | jameswan | 2024-11-01T02:29:23 | Show HN: Aux Machina – AI photo generator without complex prompting | Hey HN,<p>We’ve been on a wild journey building Aux Machina, and think it’s fun to share a bit of the ride with you. It all starts with a “what if.” What if we make AI-powered visuals without needing a PhD in prompting? Sounds straightforward, right? Turns out, it’s more like a series of sleepless nights and way more Python than we bargain for .<p>Picture this: We met at Harrison.ai in 2019, knee-deep in CNNs in a pre-ChatGPT world. We were building healthcare AI that saves lives — serious stuff. But between all the daily stand-ups and coffee refills , we realised, “Hey, we actually like working together!” (Shocking, I know.) Naturally, after life-saving AI, we took on the next obvious challenge ... helping people create pretty pictures with way less hassle . Same level of world-changing impact, right?<p>So, with a “let’s make this happen” attitude, we got to work. We’re a small, execution-focused team that just wanted to make creating custom visuals way easier for small businesses, marketers, and creators who don’t want to deal with complicated AI prompts just to get something halfway decent. “How hard could it be?” we thought. Spoiler: it’s harder than we thought. We went with a Python backend and a TypeScript/React front end, nothing too fancy, but it gets the job done (mostly without crashing) . Although, we had one glorious moment where everything went down in flames for a time ... because we didn’t update the .env file. Yup, just one tiny oversight that took hours to discover - like finding a needle in a haystack, if the haystack was also on fire. Fun times.<p>On the model side, we fine-tuned a latent diffusion model and added some low-rank adaptations to handle the image generation process. The best way to describe it? You feed in your ideas, and the model refines them through layers of noise reduction and reconstruction, bringing something (hopefully) that doesn’t look like abstract art . Sometimes it’s spot on, sometimes... not so much. But hey, we’ve all had our Picasso moments.<p>Modularity is a big deal for us. We don’t want to build one of those all-in-one, Swiss Army knife disasters that no one touches because it’s too overwhelming — and racks up tech debt faster than you can say ‘refactor’ . I’ll never forget the day we almost accidentally wipe out half the codebase because of a misplaced function. David, our solutions architect, still talks about it like a near-death experience. So yeah, we break things down into modular components. The result? We iterate quickly and improve individual features without breaking everything else — or turning the codebase into a spaghetti monster .
We also can’t resist adding some large language model (LLM) magic for tasks like interpreting user inputs or the visual vocabulary tied to reference images. The LLM makes sure what you type turns into something that’s, well, pretty close to what you’re imagining. Most of the time, anyway . There were a few days when the output was stranger than a cat riding a Roomba.<p>A few things we’ve learned on the way:
1. Building AI tools is like trying to herd cats — except the cats are on fire, and you’re trying not to get burned .
2. React is great for making things slick and modular, but if you’re not careful, it’ll eat your lunch and your weekend.
3. Latent diffusion models are super cool, but they need a lot of fine-tuning unless you want everything to look like Picasso had a rough day .<p>We’re super excited to see how people use Aux Machina as we head into the go-to-market phase . We were all sizzle and no steak, but now we’re as ready as a bride’s nightie on the big day. We’d love your thoughts (or even better, your brutally honest feedback).<p>Give it a try: <a href="https://www.auxmachina.com" rel="nofollow">https://www.auxmachina.com</a> | https://www.auxmachina.com | 5 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,531 | thunderbong | 2024-11-01T02:29:25 | How to Accelerate Code Reviews with Nudges: Insights from Microsoft's Study | null | https://reviewnudgebot.com/blog/how-to-accelerate-code-reviews-with-nudges-insights-from-microsofts-study/ | 2 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,555 | solosquad | 2024-11-01T02:35:35 | ChatGPT Search Extension | null | https://chromewebstore.google.com/detail/chatgpt-search/ejcfepkfckglbgocfkanmcdngdijcgld | 2 | 0 | null | null | null | Failed after 3 attempts. Last error: Quota exceeded for quota metric 'Generate Content API requests per minute' and limit 'GenerateContent request limit per minute for a region' of service 'generativelanguage.googleapis.com' for consumer 'project_number:854396441450'. | ChatGPT search - Chrome Web Store | null | null | OverviewChange default search engine to ChatGPT search.Note: ChatGPT search is available to all ChatGPT Plus and Team users, as well as SearchGPT waitlist users. Enterprise and Edu users will get access in the next few weeks. We’ll roll out to all Free users over the coming months.
This extension will make ChatGPT your default search engine in Chrome. Once ChatGPT has been set as your default search engine, you can search directly via your browser URL bar.
ChatGPT You can get fast, timely answers with links to relevant web sources, which you would have previously needed to go to a search engine for. This blends the benefits of a natural language interface with the value of up-to-date sports scores, news, stock quotes, and more. Chats also now include links to sources, such as news articles and blog posts, giving you a way to learn more.
In order to redirect a query to Google search, type "!g [your query]" (e.g. !g foobar) directly in your browser URL bar.
You can read more about ChatGPT search and access OpenAI's Privacy Policy here: https://help.openai.com/en/articles/9237897-chatgpt-searchTony TNov 6, 2024Not having the consideration to add both "Search ChatGPT for ...." *and* "Search Google for ...." items to the right-click context menu is a complete deal-breaker.
ChatGPT search has been disabled until this problem is addressed.14 out of 16 found this helpfulMuhammad Ali KhawajaNov 6, 2024It takes a while to show the search results. Omni bar search queries needs a snappier UI.4 out of 4 found this helpfulViníciusNov 6, 2024Amazing!!! Better than google.3 out of 5 found this helpfulDetailsVersion1.11UpdatedNovember 6, 2024Offered byOpenAISize11.81KiBLanguagesDeveloperOpenAI3180 18th St
San Francisco, CA 94110-2043
US Email [email protected] Phone +1 650-842-0962TraderThis developer has identified itself as a trader per the definition from the European Union.D-U-N-S080579740PrivacyThe developer has disclosed that it will not collect or use your data. To learn more, see the developer’s privacy policy.This developer declares that your data isNot being sold to third parties, outside of the approved use casesNot being used or transferred for purposes that are unrelated to the item's core functionalityNot being used or transferred to determine creditworthiness or for lending purposes | 2024-11-08T01:01:03 | null | train |
42,013,569 | henkxheng | 2024-11-01T02:37:51 | Show HN: Transform Any Tool into a Collaborative Space | null | https://www.remotespace.ai/ | 1 | 2 | [
42013577
] | null | null | null | null | null | null | null | null | null | train |
42,013,574 | thunderbong | 2024-11-01T02:38:38 | A Halloween scary story for dev | null | https://meowbark.dev/A-Halloween-scary-story | 1 | 0 | null | null | null | Failed after 3 attempts. Last error: Quota exceeded for quota metric 'Generate Content API requests per minute' and limit 'GenerateContent request limit per minute for a region' of service 'generativelanguage.googleapis.com' for consumer 'project_number:854396441450'. | A Halloween scary story for dev. | null | null | Happy Halloween 🎃! I have a horror story for you! Don’t forget to stock up some good candy for your kid, and feed your cat some nips while you’re at it 😎.
You’re a senior software engineer at BigTechCorp, counting your blessings since landing this role last October. In this brutal tech market, your college buddies are still sending “Hope you’re doing well!” LinkedIn messages to every recruiter they can find.
The project you’ve been grinding on for months is finally ready. Your team desperately needs this win - the last three quarters have been rough, and everyone from your manager to the VP of Engineering has been “just checking in” on your progress. You’ve been pulling extra hours, living on cold brew and takeout, but it’s worth it. This feature is going to be chef’s kiss.
It’s a perfect Friday afternoon, the kind where golden autumn sunlight streams through your home office window. The clock shows 2 PM, and you’re ready to wrap up early. Your PR has more approving comments than a celebrity’s Instagram post. Test coverage would make your CS professor proud. You take one last look at your deployment checklist - everything’s green.
What could possibly go wrong?
You hit the deploy button with the satisfied flourish of an artist signing their masterpiece. Laptop closed, you head out for a victory walk, breathing in that crisp October air. The crunch of leaves under your feet feels like nature’s applause for your achievement.
Your phone buzzes at 4:30 PM. You almost ignore it - almost. It’s your on-call teammate.
“prod is down for MegaCorp”
Your heart stops. MegaCorp isn’t just any customer - they’re the whale that pays for everyone’s free snacks and those fancy standing desks. You sprint back to your car, your peaceful walk becoming a distant memory.
Back at your desk, you take a swig of now-lukewarm coffee and fire off a quick text: “Sorry, might miss the Halloween party tonight.” Your Alien Romulus costume will have to wait - you’ve got a real monster to fight.
The local environment looks fine - mockingly fine. Every test passes with a smug green checkmark. The console is clean enough to eat off of. No API errors, no data corruption, nothing. Just the sound of your anxious keyboard tapping echoing through your empty house.
Time for the nuclear option: rolling back the deployment. In theory, it’s just one button press. In practice, it’s a prayer to the infrastructure gods. You hit revert and hover over the refresh button, watching the deployment bot’s updates scroll by in the war room channel like a digital horror movie.
Deploy complete.
Hard refresh. Harder refresh. CMD+SHIFT+R refresh - the desperate kind.
Nothing changes. The bug lives on, like a zombie 🧟 that won’t stay dead.
You step outside, gulping down air that suddenly feels too thick. Your phone is lighting up with messages from the support team. Each notification feels like another nail in your career’s coffin.
Back at your desk, you dive into the dependency rabbit hole. console.log statements multiply like tribbles. Your code starts looking like a serial killer’s manifesto, with debugger statements everywhere. Two hours later, you’re three levels deep in the internal dependency chain, somewhere no application developer was meant to go.
Then you see it. Hidden in a utility function that has a try/catch clause, deep in your dependency chain. The function was validating input data before passing it through a series of processing functions - innocent-looking code that just transforms a massive data structure. The catch block had its own processing logic, presumably to format error messages before re-throwing.
But your error was different - completely different than what the catch clause expected. The error mapping code broke because it encountered an unhandled error shape, throwing a new error with a completely different payload. The upstream code was expecting errors of a particular structure, and when it got this mutated version instead, the whole system fell like silent dominoes. A simple Cannot read property of undefined had transformed into an unrecognizable monster by the time it reached your error handlers.
Somewhere in that try clause, a function that was not supposed to fail DID fail. The kind of function that’s so fundamental, so basic, that no one even thinks to test for its failure. The type system promised it would work. The unit tests swore by it. The documentation claimed it was bulletproof.
The real kicker? The dependency had updated just an hour before your deployment, and it was very well unit tested. Like a cruel twist of fate, your feature walked right into its trap.
You fix the dependency version, verify the fix, and send out the all-clear signal at 10 PM. Your phone shows three missed calls from your friends at the Halloween party. Your Alien costume hangs on the door, judging you silently.
Monday’s going to be fun explaining how a single undefined property brought down the system. Your mind wanders to your mortgage payments, your job security, and that one tweet about tech layoffs you saw last week.
As you finally shut down your laptop, you can’t help but wonder: we have TypeScript, we have testing, we have code reviews - why is error handling still such a nightmare? Perhaps that’s the real horror story.
Next time someone suggests a Friday deployment, you’ll know better. You’ve seen what lurks in the shadows of the dependency tree, waiting to strike. Some monsters wear lab coats and carry clipboards, but the scariest ones? Undefined and error handling.
The End 🎃 | 2024-11-08T07:33:52 | null | train |
42,013,576 | mlhpdx | 2024-11-01T02:38:53 | Russia fines Google all the world’s money, and then some | null | https://www.bbc.com/news/articles/cdxvnwkl5kgo | 8 | 2 | [
42013652
] | null | null | no_error | Russia fines Google more than entire world's GDP | 2024-10-31T12:24:25.832Z | Graham Fraser | A Russian court has fined Google two undecillion roubles - a two followed by 36 zeroes - for restricting Russian state media channels on YouTube.In dollar terms that means the tech giant has been told to pay $20,000,000,000,000,000,000,000,000,000,000,000.Despite being one of the world's wealthiest companies, that is considerably more than the $2 trillion Google is worth.In fact, it is far greater than the world’s total GDP, which is estimated by the International Monetary Fund to be $110 trillion.The fine has reached such a gargantuan level because - as state news agency Tass has highlighted - it is rapidly increasing all the time.According to Tass, Kremlin spokesman Dmitry Peskov admitted he "cannot even pronounce this number" but urged "Google management to pay attention."The company has not commented publicly or responded to a BBC request for a statement.Russia media outlet RBC reports the fine on Google relates to the restriction of content of 17 Russian media channels on YouTube.While this started in 2020, it escalated after Russia's full-scale invasion of Ukraine two years later.That saw most Western companies pull out of Russia, with doing business there also tightly restricted by sanctions.Russian media outlets were also banned in Europe - prompting retaliatory measures from Moscow.In 2022, Google’s local subsidiary was declared bankrupt and the company has stopped offering its commercial services in Russia, such as advertising.However, its products are not completely banned in the country.This development is the latest escalation between Russia and the US tech giant.In May, 2021, Russia’s media regulator Roskomnadzor accused Google of restricting YouTube access to Russian media outlets, including RT and Sputnik, and supporting "illegal protest activity".Then, in July, 2022, Russia fined Google 21.1bn rouble (£301m) for failing to restrict access to what it called "prohibited" material about the war in Ukraine and other content.There is virtually no press freedom in Russia, with independent news outlets and freedom of expression severely curtailed. | 2024-11-08T16:16:16 | en | train |
42,013,587 | kamaraju | 2024-11-01T02:41:04 | ChatGPT Search | null | https://chatgpt.com/?hints=search | 1 | 1 | [
42013646
] | null | null | null | null | null | null | null | null | null | train |
42,013,589 | lxm | 2024-11-01T02:41:17 | Kraken Lays Off 15% of Workers | null | https://www.nytimes.com/2024/10/30/technology/kraken-cryptocurrency-layoff-ceo.html | 9 | 0 | null | null | null | null | null | null | null | null | null | null | train |
42,013,591 | betap | 2024-11-01T02:41:56 | A Lost Mayan City Has Been Found with Laser Mapping | null | https://www.wired.com/story/lost-mayan-city-valeriana-lidar-discovery/ | 2 | 0 | null | null | null | no_error | A Lost Mayan City Has Been Found With Laser Mapping | 2024-10-30T06:00:00.000-04:00 | Anna Lagos | A Mayan city lost in the dense jungle of southern Mexico has been revealed. The discovery occurred in the southeastern state of Campeche, and archaeologists have named it Valeriana, after a nearby freshwater lagoon.“The larger of Valeriana's two monumental precincts has all the hallmarks of a classic Mayan political capital: enclosed plazas connected by a broad causeway; temple pyramids; a ball court; a reservoir formed by damming an arroyo (a seasonal watercourse); and a probable E-Group assemblage, an architectural arrangement that generally indicates a founding date prior to AD 150,” says the study, published in the journal Antiquity.The city's discovery didn't require breaking through the jungle with machetes or patiently excavating with brushes and spatulas. Nor did researchers need tape measures, binoculars, or compasses to find their way through the thick foliage. Instead, they employed state-of-the-art technology: lasers, drones, and satellite maps. With these tools, they discovered a city hidden for centuries beneath the thick Mexican jungle, unearthing pyramids, enclosed plazas, and an ancient reservoir.Luke Auld-Thomas, a PhD student at Tulane University, made the discovery. His analysis revealed a huge network of previously unexplored settlements.Auld-Thomas and his fellow researchers have succeeded in mapping the city beneath the jungle thanks to airborne laser scanning, better known as lidar (light detection and ranging), a remote-sensing technique that uses pulsed lasers and other data collected through flyovers that can generate accurate three-dimensional models of surface features, revolutionizing the way archaeologists explore the hidden past.Laser pulses generate a topographic map in a manner similar to how a bat uses echolocation: Laser light is fired from an aircraft, bounces off objects on the ground, and returns to the detector located on the underside of the aircraft. In Mexico, although only a small fraction of the pulses pass through the dense jungle, the large number of pulses emitted allows enough light to reach the ground, creating a map with a resolution of up to 1 meter. Based on the timing and intensity of the returning pulses, the detector can map the contours of the terrain, revealing hills, ditches, and ancient ruins covered in vegetation. The technology is also being integrated into autonomous cars to help them avoid crashes.“For a long time, our understanding of the Mayan civilization was limited to an area of a few hundred square kilometers,” Auld-Thomas says. “This limited sample was obtained with great effort, with archaeologists painstakingly scouring every square meter, hacking away at vegetation with machetes, only to discover they were standing on a pile of rocks that might have been someone's house 1,500 years ago.”While Auld-Thomas knew that lidar could be a valuable tool, he was also aware of its high cost. Funders are often reluctant to invest in lidar surveys in areas where there is no visible evidence of Mayan settlement, despite the fact that this civilization reached its peak between 250 and 900 AD.Campeche: A Center of Dense Urbanization Since the Mayan EraIn this case, the lidar data was originally collected over a decade ago, for completely unrelated purposes. The scans were completely in 2013 by the Mexican firm CartoData, using a Riegl LMS-Q780 sensor. Processing was carried out by the Woods Hole Research Center (WHRC), and the data was made publicly available a few years later by the M-REDD+ Alliance.The dataset includes three transects and three study blocks. The transects have an approximate width of 275 meters and a total length of 213 kilometers, covering an area of 58.3 square kilometers. The survey blocks cover a total area of 64.1 square kilometers, distributed in three locations: south of the town of Xpuhil, near the archaeological site of Río Bec; near the villages of Dos Lagunas and Bel Ha; and near the town of Ucum, in northern Campeche.The study mentions that the analysis of 6,764 structures in the lidar data blocks reveals a settlement density of 55.3 structures per square kilometer, comparable to other research in the region. These data are useful for assessing settlement density on a regional scale and exceed values recorded in Belize and Guatemala. However, they do not provide a complete picture of the level of urbanization, which requires analysis of local variability and density gradients. For this purpose, a kernel density estimation was applied to the study blocks, the results of which are consistent with the densities recorded in other Mayan archeological sites such as Oxpemul and Becan.Archeologists in the 20th century were correct in stating that the interior of Campeche is a substantially anthropogenic landscape, i.e., human-modified, with urbanized areas where rural populations interacted with dense cities. Settlement density data, ranging from 49 to 61 structures per square kilometer, indicate that cities and dense settlements are common in large parts of the central Maya lowlands. New discoveries, such as the city of Valeriana, reinforce this view, showing that urbanization was a widespread phenomenon in the region.Lidar images produced by the study's authors.
Courtesy of Luke Auld-Thomas et al/Cambridge University PressArchaeologists increasingly recognize that the world's tropics and subtropics hosted a wide variety of urban forms in antiquity. Many of these settlements followed a pattern of spatial dispersion, commonly called “low-density urbanism.” However, it is now being recognized that these urban landscapes were not uniform, but exhibited significant variations in settlement density, both within and around cities and between subregions.At the same time, the growing body of research has revealed a greater abundance of settlements and cities than had previously been contemplated. This has generated a tension between two developments: On the one hand, the recognition of high variability in settlement density and, on the other, evidence of a more densely urbanized past than previously thought.Although lidar was developed in the 1960s to study clouds and atmospheric particles, its application in archaeology is relatively recent. It was not until the last decade that archaeologists began employing it to unearth hidden landscapes. In 2009, archaeologists Diane and Arlen Chase of the University of Nevada, Las Vegas, pioneered the use of lidar to map a Mayan city, revolutionizing the way ancient civilizations are detected and studied.According to the study, some researchers argue that the discovered landscapes reflect a high population density, while others suggest that the surveys are biased and overrepresent the most densely populated areas. This leaves open the question of whether as yet unexplored areas could confirm the existence of a higher urban density or show less dense occupation.This story originally appeared on WIRED en Español and has been translated from Spanish. | 2024-11-08T06:41:59 | en | train |
42,013,600 | xanderlewis | 2024-11-01T02:45:18 | Writing Mathematics in Pencil, and Why Analogue Is Not Dead (2017) | null | https://nhigham.com/2017/02/21/writing-mathematics-in-pencil-and-why-analogue-is-not-dead/ | 3 | 2 | [
42013679
] | null | null | Failed after 3 attempts. Last error: Quota exceeded for quota metric 'Generate Content API requests per minute' and limit 'GenerateContent request limit per minute for a region' of service 'generativelanguage.googleapis.com' for consumer 'project_number:854396441450'. | Writing Mathematics in Pencil, and Why Analogue is Not Dead | 2017-02-21T12:19:00+00:00 | null |
It’s an old joke that mathematicians need just a pencil, paper, and a bin, while philosophers are even more frugal because they don’t need the bin. Yet nowadays more and more of the time of mathematicians, indeed all scientists, is spent at the computer. Whereas twenty years ago I would handwrite a draft of a paper before typing it in, I now do almost all the drafting directly in at the keyboard.
But in response to computers dominating our lives, and in a move away from the (mythical?) paperless office, people are increasingly reverting to analogue tools, encouraged by the pleasure of handling stationery and, for those of us who were brought up in an analogue world, nostalgia.
This is a good time to employ retro tools in a digital world because we can now buy online an increasingly wide variety of stationery from all around the world.
What might you gain by writing mathematics with a pencil and paper as opposed to typing it at the computer? Sitting at a desk with a pencil in hand you are free from the distractions of the windows on your computer screen. The analogue process, with its delays of turning a page, sharpening the pencil, and rubbing out mistakes, has the benefit of slowing you down and thereby promoting your flow of thought and creativity. And the touch of the paper and the smell of cedar as you sharpen the pencil refresh the senses.
Donald Knuth has another reason for writing with a pencil, as explained in this 2008 interview.
I love keyboards, but my manuscripts are always handwritten. The reason is that I type faster than I think. There’s a synchronization problem. I can think of ideas at about the rate I can write them down with a pencil. But with typing I’m going faster, so I have to sync, and my thoughts have to start up and stop again in a way that involves more of my brain.
Yet more reasons for using pencils are given in the video Why Use Pencils? by T. J. Cosgrove. One of T. J.’s points is that the graphite produced by a pencil does not fade, unlike inks.
It is also worth noting that in some recently published research psychologists found evidence that students who take notes with pencil (or pen) and paper outperform those who take notes on a laptop.
So there are some good reasons for writing with a pencil. How should you choose one from among the many different types available?
I don’t know of a good source of advice on pencils for mathematicians (maybe I will write something in due course), but this blog post on pencils for musicians is largely applicable if you replace “music” by “mathematics”. The post is by Caitlin Elgin, from the wonderful Manhattan pencil shop pictured in the photo above, which I took when I visited it last year.
Post navigation
| 2024-11-07T22:47:48 | null | train |
42,013,601 | julio1909x | 2024-11-01T02:45:36 | null | null | null | 1 | null | null | null | true | null | null | null | null | null | null | null | train |
42,013,608 | lxm | 2024-11-01T02:47:46 | Sugar Rationing Lowered Diabetes and Hypertension in British Children | null | https://www.nytimes.com/2024/10/31/health/sugar-rationing-britain-health-experiment.html | 31 | 13 | [
42013719,
42013796,
42014016
] | null | null | 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.