Explanation of everything you can see in htop/top on Linux

49 theanonymousone 2026-07-04 12:30:07 UTC peteris.rocks Llama Share 7

The load average of 1.0 on a two-core machine does not mean 50% CPU usage, but rather 100% CPU utilization, as the load average is the average number of running processes, not CPU usage. The load average is calculated by counting the number of running and uninterruptible processes, and is an exponentially damped moving average of the system load since startup.

Potential session/cache leakage between workspace instances or consumer accounts

10 chatmasta 2026-07-04 14:05:07 UTC github.com/anthropics Full Share 2

Bug Description Apparent session leakage, despite authenticated to Enterprise ZDR workspace. Agent suddenly started asking me what kind of bricks I wanted for my Minecraft temple and confidently asserted in its recap that it's building a...

Astrophysicists Puzzle over Webb’s New Universe

101 jnord 2026-07-04 09:30:07 UTC quantamagazine.org Llama Share 50

Astrophysicist Charlotte Mason and colleagues are studying "little red dots" discovered by the James Webb Space Telescope, which may be black holes or a new type of object. Researchers are exploring various theories to explain the dots' existence and the universe's early mysteries.

The Vespa at 80: Why the Italian scooter remains the coolest thing on 2 wheels

61 cf100clunk 2026-07-04 10:30:07 UTC cbc.ca Llama Share 53

A Vespa museum in Rome showcases the scooter's 80-year history, from its birth in 1946 to its current status as a symbol of freedom and style. The Vespa's unlikely story began as a solution to post-war Italy's transportation needs, designed by Corradino D'Ascanio and launched by Enrico Piaggio.

Maybe you should learn something

229 tylerdane 2026-07-04 04:30:07 UTC marginalia.nu Llama Share 111

Learning new things is fun and makes life richer, but it requires a daily time investment of up to an hour. It's a long journey with initial struggles, but consistent practice leads to improvements over time.

Postgres data stored in Parquet on S3: LTAP architecture explained

81 andrenotgiant 2026-07-04 06:30:07 UTC databricks.com Llama Share 24

The author, a co-founder of Databricks, discusses how traditional OLTP databases are clunky, difficult to scale, and fragile, and proposes a new architecture called Lakebase that externalizes storage and compute, making databases more durable, elastic, and cheaper to run. Lakebase's design also enables a new feature called LTAP, which unifies transactional and analytical workloads by storing ...

Night Witches – all-female Soviet aviator regiment WW2

33 gverrilla 2026-07-04 11:30:07 UTC en.wikipedia.org Llama Share 7

The 588th Night Bomber Regiment, known as the "Night Witches," was an all-female Soviet military unit that flew over 23,000 sorties during World War II. The regiment was formed in 1941 and flew harassment and precision bombing missions against the German military until the end of the war in 1945.

The bottleneck might be the air in the room

508 gslin 2026-07-04 07:30:07 UTC blog.mikebowler.ca Llama Share 311

High-stakes meetings in poorly ventilated rooms can impair decision-making due to high CO2 levels. A CO2 monitor can help identify and fix this issue, which is often overlooked and can be as simple as opening a window.

Performance per dollar is getting faster and cheaper

298 latchkey 2026-07-03 22:30:07 UTC wafer.ai Llama Share 110

AMD's Instinct MI350 series offers cheaper GPUs with comparable specs to NVIDIA's Blackwells, but lacks day-0 support for frontier models. With optimization and tweaks, AMD can achieve competitive performance per dollar, closing the gap with NVIDIA, but still requires more support and development.

Leanstral 1.5: Proof abundance for all

296 programLyrique 2026-07-03 23:30:07 UTC mistral.ai Llama Share 86

Leanstral 1.5 is a free, open-source model that delivers a performance upgrade for formal verification, achieving a new state-of-the-art on FATE-H/X and PutnamBench. It verifies complex code properties and uncovers previously unknown bugs in open-source repositories.

Costco is the anti-Amazon

464 bookofjoe 2026-07-03 15:30:07 UTC phenomenalworld.org Llama Share 423

Amazon's complex logistics model may not be the most efficient or socially beneficial, while Costco's simpler model offers lower costs and better worker treatment. A public grocery store model, inspired by Costco and the military commissary system, could provide cheaper groceries for New Yorkers with lower overhead.

Mir Books – Books from the Soviet Era

122 clmul 2026-07-04 04:30:07 UTC mirtitles.org Llama Share 52

The book is a narrative by geologist Ekaterina Radkevich about her expeditions and research on Earth and its resources. It completes the Science for Everyone Series, a collection of scientific knowledge for the general public.

Giant trees have no trouble pumping water to top branches: new research

232 hhs 2026-07-03 23:30:07 UTC news.exeter.ac.uk Llama Share 103

New research reveals that the world's tallest tropical trees can pump water to their topmost branches without difficulty. The study found that these trees' water systems are perfectly adapted to their height, making them no more vulnerable to drought than shorter trees.

Steam Controller Auto-Charge – pilot to magnetic charging puck using CV

166 zdw 2026-07-03 23:30:07 UTC github.com/fossprime Llama Share 36

Steam Controller Auto-Charge is an open-source app that automatically charges a Steam Controller using computer vision and WebHID telemetry. It uses optical flow tracking, haptic navigation, and battery status polling to guide the controller to its charging puck.

Agentic coding notes from Galapagos Island

135 gm678 2026-07-04 05:30:07 UTC danluu.com Llama Share 63

The user discusses their experience with AI and testing, highlighting the benefits of using AI to generate tests and the limitations of relying solely on LLMs for testing. They propose a testing methodology that combines fuzzing, randomized testing, and human review to achieve high-quality results, and discuss the importance of feedback loops and independent perspectives in reducing false positives.

Jamesob's guide to running SOTA LLMs locally

374 livestyle 2026-07-03 15:30:07 UTC github.com/jamesob Llama Share 168

User has $2k to spend on local machine intelligence and $40k for top-of-the-line. They have 4x RTX Pro 6000s and built a DDR4 system to host them.

MSI Center – How to gain SYSTEM privileges in seconds

115 MrBruh 2026-07-04 01:30:07 UTC mrbruh.com Llama Share 48

The user discovered a severe vulnerability in MSI Center, a pre-installed software on MSI laptops and desktops, allowing for Remote Code Execution (RCE) with valid login credentials. MSI patched the vulnerability within two days and the user is still waiting for a CVE from VulDB after submitting their report.

Synthesis is harder than analysis

117 azhenley 2026-07-04 03:30:07 UTC surfingcomplexity.blog Llama Share 30

Calculus has two main branches: differential calculus, which calculates the slope of a function, and integral calculus, which calculates the area under a curve. The two are connected by the Fundamental Theorem of Calculus, which states that integrals are anti-derivatives, but integral calculus is generally harder than differential calculus due to its global nature.

FreeBSD ate my RAM

165 theanonymousone 2026-07-03 19:30:07 UTC crocidb.com Llama Share 66

The user investigated why memory usage reported by fastfetch and btop differed on FreeBSD, discovering that the discrepancy was due to btop's incorrect handling of the ARC cache and filesystem metadata buffer cache. The user contributed patches to btop and htop to improve their memory usage reporting, and also submitted a PR to fastfetch, which was later merged with improvements.

Breaking the Bird Barrier: Scientist Decodes Zebra Finch Language

4 yyyk 2026-07-04 13:30:07 UTC freepressjournal.in Prefix Share 0

Dr Julie Elie of the University of California, Berkeley, has won the 2026 Coller-Dolittle Prize for Two-Way Interspecies Communication, receiving $100,000 for decoding the core vocabulary of zebra finches and advancing research into human-animal communication. Elie identified the 11 core calls used by zebra finches and decoded their meanings. Her research showed that the birds use distinct ...

SearXNG: A free internet metasearch engine

243 theanonymousone 2026-07-03 20:30:07 UTC github.com/searxng Full Share 65

SearXNG is a free internet metasearch engine which aggregates results from various search services and databases. Users are neither tracked nor profiled. - searxng/searxng

The firefighting system of the Van der Heyden brothers in 17th century Amsterdam

113 zdw 2026-07-03 23:30:07 UTC worksinprogress.co Llama Share 20

Amsterdam's fire-fighting system improved in the 17th century with the introduction of new engines and a culture of inventiveness. The city's fire damage decreased dramatically after implementing the Van der Heyden brothers' suction and pressure hoses.

2026 Unslop AI-Written Fiction Contest Results

35 networked 2026-07-04 06:30:07 UTC hyperstitionai.com Llama Share 96

The 2026 Hyperstition Unslop AI fiction writing contest awarded $10,000 to A. Best for his story "The June" and highlighted the emergence of AI allegory steganography in many submissions. Judges noted that while some stories showed promise, they often fell into predictable patterns and lacked the diversity and creativity of human-written fiction.

Odin, Wikipedia and engagement farming

197 stock_toaster 2026-07-03 23:30:07 UTC katamari64.se Llama Share 266

GingerBill, the creator of the Odin programming language, expressed concerns about Wikipedia's deletion of the Odin article due to lack of reliable sources, but his tone shifted from dismissive to concerned after Casey Muratori, a famous videogame programmer, pointed out the flaws in Wikipedia's notability rules for programming languages. Casey argued that Wikipedia's rules prioritize ...

Ship traces journey Spanish Armada sailors made in 1588

17 austinallegro 2026-07-04 10:30:07 UTC irishtimes.com Llama Share 8

The Spanish galleon Galeón Andalucía sailed along Ireland's west coast, visiting Streedagh Beach in Co Sligo where three Armada ships sank in 1588. The ship's visit is part of the Remembering the Armada Festival, boosting tourism and commemorating the disaster.

New serious vulnerabilities spiked around release of Claude Mythos Preview

132 cubefox 2026-07-03 21:30:07 UTC epoch.ai Llama Share 57

Anthropic and OpenAI announced efforts to use frontier models for cybersecurity, but this led to a significant increase in high-severity vulnerabilities. The number of high- and critical-severity CVEs jumped 3.5x in June after the announcements.

The Scanline Sweeper: A Glyph Rendering Algorithm [pdf]

25 kouosi 2026-07-04 08:30:07 UTC rookandpossum.com Llama Share 2

The Scanline Sweeper is a GPU-based Bézier curve rasterization algorithm that estimates coverage analytically without computing explicit winding numbers or tessellation, resulting in a numerically robust algorithm. It operates on glyph curves directly on the GPU, integrating signed areas swept by each curve intersecting the scanline of interest to produce an anti-aliased result for a single ...

Gone but Not Forgotten: Recovering the Dead Web

87 wslh 2026-07-03 23:30:07 UTC blog.archive.org Llama Share 32

The Pew Research Center study found that 38% of webpages from 2013 are no longer accessible, but the Wayback Machine has rescued roughly 15% of those dead pages. Web archives like the Wayback Machine play an increasingly important role in rescuing the dead Web and minimizing the fracture of the knowledge graph on the Web.

Show HN: Classify mechanical faults using Contrastive Language-Audio Pretraining

29 dataviz1000 2026-07-04 04:30:07 UTC github.com/adam-s Llama Share 2

cardiag is a proof-of-concept audio-ML pipeline for car fault diagnosis from phone recordings. It provides a calibrated triage aid with honest uncertainty, isolating mechanical sounds and ranking likely parts.

Soatok's Informal Guide to Threat Models

112 zdw 2026-07-04 01:30:07 UTC soatok.blog Llama Share 27

The user discusses the importance of threat modeling in cybersecurity, explaining that it's a formal process that helps identify potential threats and vulnerabilities in a system, and provides a framework for understanding assumptions and accepted risks. They also apply threat modeling to real-world examples, including hybrid post-quantum cryptography and end-to-end encrypted messaging apps, ...