HackerNews Digest

April 11, 2026

Filing the corners off my MacBooks

The author documents a personal modification of MacBook aluminum unibody cases by filing down the sharp corners, especially around the notch, to improve wrist comfort. The post notes that Apple’s design creates a very acute edge, which the author mitigates using a rough file followed by progressive sanding with 150‑ and 400‑grit paper. Precautions taken include taping over the speakers and keyboard, clamping the laptop to a workbench, and filing in incremental steps to avoid penetrating the chassis. The resulting surface is described as smooth, with later photographs showing typical wear over time. The author states this alteration was performed on a work machine, intends to repeat it on future devices, and offers assistance to others interested in similar modifications.
Read full article →
Comments cluster around discomfort with Apple laptops’ sharp aluminum edges, prompting many users to file, sand, or otherwise blunt them for ergonomic relief. A sizable portion praises the DIY modifications as satisfying personalization and a “right‑to‑repair” stance, while others caution about warranty, resale value, or potential structural impacts. Opinions on the original design are split: some criticize it as needlessly aggressive and poorly considered, whereas a minority find the aesthetic appealing or merely humorous. Overall, the community favors practical edge‑softening despite lingering concerns about risks and design intent.
Read all comments →

Artemis II safely splashes down

Reentry for Artemis II is scheduled to begin at 7:53 p.m. ET. Twenty‑four seconds after entry, at roughly 1,950 mi from the splashdown zone, plasma generated by the 16.5‑ft heat shield will block communications. One minute 22 seconds into descent the shield temperature peaks near 5,000 °F (≈½ the Sun’s surface temperature). The blackout lasts about six minutes, after which the capsule is traveling ~9,000 mph and is <200 mi from the target point. At ~8 min post‑entry the vehicle reaches ~100,000 ft altitude; one minute later it falls below Mach 1. At ~50,000 ft and ~300 mph, 11 parachutes deploy sequentially, followed by three pilot chutes that extract three 116‑ft main chutes at ~6,000 ft. These reduce the descent rate to ~17 mph for a splashdown in the Pacific Ocean near San Diego.
Read full article →
The comments express strong enthusiasm and pride in the successful Artemis II splashdown, highlighting the mission’s inspirational impact, the technical achievement of precise re‑entry and heat‑shield performance, and a sense of national and personal identity tied to space exploration. Simultaneously, contributors note the heightened risk compared with earlier programs, critique the political framing and perceived propaganda, and point out minor technical or communication issues observed during the broadcast. Overall, the reaction combines celebratory admiration with measured skepticism about risk, messaging and future direction.
Read all comments →

1D Chess

1D‑Chess is a linear chess variant featuring only three piece types—King, Knight, and Rook—played on a single‑dimensional board. The King moves one square in either direction; the Knight moves exactly two squares forward or backward, jumping over intervening pieces; the Rook moves any number of squares in a straight line in either direction. Victory is achieved by checkmating the opponent’s king, i.e., delivering check with no legal escape. Draws occur by stalemate (king not in check with no moves), three‑fold repetition, or insufficient material (only kings remaining). The game can be played against an AI, and a sample forced‑win line for White is given: N4 N5, N6 K7, R4 K6, R2 K7, R5++. The variant was originally described by Martin Gardner in the July 1980 issue of *Scientific American* (Mathematical Games column).
Read full article →
The comments show widespread enthusiasm for the one‑dimensional chess variant, with many participants describing the puzzles as enjoyable, intuitive, and surprisingly engaging. Several users discuss specific mechanics such as knight and rook behavior, stalemate interpretation, and zugzwang, often noting moments of confusion that were later resolved through hints or trial‑and‑error. A minority express frustration over unclear notation or perceived difficulty, while others draw parallels to other games and concepts, adding humorous or tangential remarks. Overall, the reaction is largely positive, highlighting both the novelty and the learning curve of the variant.
Read all comments →

Chimpanzees in Uganda locked in eight-year 'civil war', say researchers

Researchers led by anthropologist Aaron Sandel of the University of Texas, co‑director of the Ngogo Chimpanzee Project, report that chimpanzee communities in Uganda are engaged in intense, territorial conflicts described as a “civil war.” The study documents frequent hostile encounters between neighboring groups, including aggressive displays, fights, and lethal attacks. These interactions underscore the species’ strong territoriality and suggest that inter‑group competition can drive sustained violence comparable to human warfare. Field observations detail the patterns of aggression, group boundaries, and the social dynamics that fuel the conflict, providing new insights into chimpanzee behavior and the ecological pressures shaping their societies.
Read full article →
The discussion centers on the chimpanzee conflict following a disease‑driven loss of individuals, interpreting it through evolutionary and game‑theoretic lenses that link resource limits, social network size, and the death of key connectors to group destabilization. Participants draw parallels to human tribalism, citing concepts such as Dunbar’s number, the role of “us‑vs‑them” bias, and historical anthropological debates, while also contrasting chimp aggression with bonobo pacifism. Overall, the tone is analytical and inquisitive, expressing curiosity about the mechanisms behind inter‑group violence and their relevance to human societies.
Read all comments →

Installing Every* Firefox Extension

The author scraped the entire Mozilla Add‑ons API to collect every public Firefox extension. Using the `/api/v5/addons/search/` endpoint with various sort parameters (recommended, users, created, updated, downloads) and the `exclude_addons` filter, they fetched all pages in parallel, ultimately obtaining 84 235 unique extensions—about 99.94 % of the ~84 000 listed. The dataset (≈49 GB, avg 585 kB per extension) is hosted on Hugging Face as `all_extensions.json`. Analysis shows the largest extensions range from 115 MB to 196 MB (e.g., dmitlichess, ReactBot Web, AI‑based tools) while the smallest is a 7.5 kB tab‑saver. The lowest‑rated extension (≥10 ratings) is “Tab Stack for Firefox.” The earliest extension is “Web Developer.” The most‑screenshot‑rich add‑on is RDS Bar (54 screenshots). The most prolific developer, “Dr. Bis,” has 84 published extensions. The author identified phishing extensions (homoglyph attacks, clipboard theft) and SEO‑spam extensions with fabricated homepage links, as well as a large group of “Custom Web Search” PUAs linked to Innover Online Group Ltd, collectively accounting for >700 k users.
Read full article →
Comments express strong amusement at the chaotic overload of Firefox extensions, describing the presentation as funny and entertaining. Reviewers note the absurd scale of thousands of extensions causing page failures and highlight a discovered performance bug linked to the extensions.json handling. Some compare the situation to outdated browser toolbar overloads and suggest similar behavior in Chrome. While a few lament the inability to load pages, the predominant tone is light‑hearted appreciation of the writing and the comedic illustration of extension bloat.
Read all comments →

Investigating Split Locks on x86-64

Split locks are atomic operations that span two cache lines, causing CPUs to fall back to a bus‑lock‑like mechanism because they cannot lock multiple lines simultaneously. This dramatically increases latency and reduces memory bandwidth, especially when accesses miss L1/L2 caches. Benchmarks using InterlockedCompareExchange64 on various x86‑64 CPUs show split‑lock latencies ranging from ~500 ns on Zen 5 to >7 µs on Intel Arrow Lake, with severe performance regressions (≈10×) for L2/L3 accesses and DRAM bandwidth cuts. Architectures differ: AMD Piledriver and Intel Goldmont Plus exhibit relatively modest penalties, while modern Intel (Arrow Lake, Alder Lake) and AMD Zen 2/Zen 5 suffer large slowdowns, though Alder Lake’s memory subsystem isolates other workloads. Linux traps split locks and adds millisecond‑level delays to mitigate noisy‑neighbor effects, which the author argues is appropriate for servers but overkill for desktops. Overall, developers should avoid split‑lock usage, and hardware vendors need clearer documentation and better handling of these operations.
Read full article →
The comment expresses confusion about the rationale for using atomic operations that are not naturally aligned, specifically those that span multiple cache lines, and questions why such split locks have been employed in games without causing issues on AMD Zen 2 and Zen 5 processors. It seeks clarification on the underlying need for this approach, indicating uncertainty about its benefits or necessity in contemporary software design.
Read all comments →

WireGuard makes new Windows release following Microsoft signing resolution

WireGuardNT v0.11 and WireGuard for Windows v0.6 have been released, updating both the low‑level kernel driver/API (WireGuardNT) and the higher‑level management UI, CLI tools, and utilities (WireGuard for Windows). New capabilities include removal of individual allowed IPs without dropping traffic and support for very low IPv4 MTUs. The release focuses on extensive bug fixes, performance gains, and code simplification achieved by raising the minimum supported Windows version, eliminating legacy compatibility layers. Toolchain components have been upgraded: a newer EWDK for the driver, updated Clang/LLVM/MingW for userspace, a recent Go release for the UI, and refreshed EV code‑signing infrastructure. The built‑in updater will prompt users to install the update; the installer is also available at https://download.wireguard.com/windows-client/wireguard-installer.exe. Testing has been performed on Windows 10 build 10240 (the oldest supported version). A prior Microsoft signing‑certificate block was resolved after brief bureaucratic delay, allowing the release to proceed.
Read full article →
The discussion centers on Microsoft’s recent developer‑account deactivations that impacted projects such as WireGuard, VeraCrypt and others. Commenters express relief that the specific cases were reversed quickly, but they also criticize the opaque verification process, inadequate communication, and the broader code‑signing mandate that they view as a systemic risk to open‑source software. The prevailing view attributes the problem to bureaucratic incompetence rather than deliberate malice, while many call for clearer policies and safeguards to prevent similar disruptions for smaller developers.
Read all comments →

Industrial design files for Keychron keyboards and mice

The repository supplies production‑grade CAD assets for Keychron keyboards and mice, covering 88 device models and more than 686 files in STEP, DXF, DWG, and PDF formats. Files are organized by series (C‑Pro, Q, Q‑HE, Q‑Pro, Q‑Max, Q‑Ultra‑8K, K‑Pro, K‑Max, K‑HE, V‑Max, P‑HE, L, and M series mice) and include plates, cases, full models, stabilizers, and keycap profiles. The source‑available license permits personal, educational, and commercial use for compatible accessories, while prohibiting copying or selling Keychron products and use of Keychron trademarks. Documentation provides a file‑format guide, getting‑started guide, 3‑D‑printing guide, and scripts for inventory generation. The repo encourages community contributions (bug fixes, ISO layouts, documentation) and offers a Discord for collaboration. Recent updates (April 2026) added K0 Max, Q12 HE, Q6 Max, K10 HE, and mouse designs. The project aims to lower entry barriers for hobbyists and engineers, enable hardware remixing, and support an ecosystem of compatible add‑ons.
Read full article →
Comments show strong enthusiasm for Keychron products, especially the Q‑series and K models, with many users highlighting solid build quality, switch replaceability, and the appeal of open‑source design files. Praise is balanced by practical criticisms: heavy chassis, inconsistent keycaps, limited battery life, slow charging, and insufficient documentation or licensing clarity. Several participants note the lack of physical retail venues for trying keyboards and express interest in ergonomic or split layouts. Overall sentiment is generally positive toward the hardware and open‑source approach, but users seek better support, clearer licenses, and refined ergonomics.
Read all comments →

AI assistance when contributing to the Linux kernel

The document outlines how AI tools may assist in Linux kernel development while adhering to existing processes and licensing rules. Contributions must follow the standard kernel development workflow, including the development‑process, coding‑style, and patch‑submission guidelines. All code must be GPL‑2.0‑only compatible and use proper SPDX identifiers, as detailed in the license‑rules documentation. AI agents are prohibited from adding “Signed‑off‑by” tags; only human submitters can certify the Developer Certificate of Origin (DCO) and must review any AI‑generated code, ensure licensing compliance, add their own Signed‑off‑by, and assume full responsibility. When AI assistance is used, an “Assisted‑by” tag should be added, formatted as Assisted‑by: AGENT_NAME:MODEL_VERSION [TOOL1] [TOOL2], where optional tools may include analysis utilities such as coccinelle, sparse, smatch, or clang‑tidy. Basic development tools are excluded from this tag.
Read full article →
The comments acknowledge the policy’s intent to keep human contributors legally responsible for AI‑generated code, viewing it as a sensible, common‑sense approach. However, many express doubts about practical enforceability, especially regarding GPL compliance and detecting license violations in model‑produced output. Several contributors worry that the rule may shift liability without effectively preventing infringement, and some view the reliance on AI as a threat to open‑source integrity. Overall, the discussion balances acceptance of responsibility with significant skepticism about feasibility and potential risks.
Read all comments →

Bevy game development tutorials and in-depth resources

The site offers a series of Bevy game‑development guides originally created for a private learning group and now expanded for public use. Authored by a programmer with a decade of Ruby web experience, the site itself is generated with a custom Ruby‑based static site generator called Staticky. It includes links to additional resources such as a Bevy starter project, a curated collection of notable Bevy repositories, and a Rust‑on‑Solana blog (Soldev). All tutorials target Bevy version 0.18. New users are directed to the “Pong Tutorial” for introductory learning, while more experienced developers are encouraged to start with the concise “TLDR” guide.
Read full article →
Comments note a scarcity of free, comprehensive resources for the Bevy engine, observing that most available material is paid. There is also surprise at the transition from Ruby to Rust, emphasizing the perceived difficulty of moving between the two languages and suggesting that Rust presents a distinctly different development experience. Overall, the feedback highlights limited accessible learning options and curiosity about the challenges associated with adopting Rust for game development.
Read all comments →