<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:media="http://search.yahoo.com/mrss/"><channel>
<title>TRMNL Leaderboard: e-ink &amp; dashboards news</title>
<link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink</link>
<description>E-ink screens, e-readers, DIY projects and home dashboards: complete articles written from specialized sources.</description>
<language>en</language>
<atom:link href="https://nbbou81000.github.io/trmnl-leaderboard/feed-eink-en.xml" rel="self" type="application/rss+xml"/>
<item><title>Farcall: An IBM PC 1981 Emulator Accessible via SSH</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/5b442e95402b</link><guid isPermaLink="false">eink-5b442e95402b-en</guid><pubDate>Fri, 20 Nov 2026 08:10:36 GMT</pubDate><category>maker</category><media:content url="https://korben.info/farcall-emulateur-dos-ssh/farcall-emulateur-dos-ssh-1.png" medium="image"/><description><![CDATA[A Rust-based emulator recreates the first IBM PC (8088) on a public server, accessible via SSH with retro games included.]]></description><content:encoded><![CDATA[<p><img src="https://korben.info/farcall-emulateur-dos-ssh/farcall-emulateur-dos-ssh-1.png" alt=""></p><p><strong>A Rust-based emulator recreates the first IBM PC (8088) on a public server, accessible via SSH with retro games included.</strong></p><p>Farcall is a Rust-based emulator designed to recreate the first IBM PC released in 1981, featuring an 8088 processor. Hosted on a public server, it can be accessed via a straightforward SSH connection. Once logged in, users enter a functional DOS environment with several retro games and tools: <strong>Zork</strong>, <strong>NetHack</strong>, <strong>Sopwith</strong>, <strong>Digger Remastered</strong>, and <strong>GW-BASIC</strong>.</p><p>Access is granted using the command ssh me@intodos.com. The system relies on <strong>GLaBIOS</strong> (an open-source BIOS) and <strong>FreeDOS</strong>, with the <strong>SvarCOM</strong> shell. For proper graphics display, a terminal supporting the <strong>Kitty graphics</strong> protocol (such as Kitty, Ghostty, or iTerm2) is required. The launcher allows users to virtually insert floppy disks, run programs, and return to the menu with <strong>Ctrl-]</strong>.</p><p>The author built Farcall by manually implementing each 8088 instruction, starting from a simple "Hello World" program. The project also includes emulation of hardware components required by the BIOS, such as the DMA controller or PIT timer. The code is released under the MIT license on GitHub and can be compiled locally using <strong>Cargo</strong>, then executed with floppy disk images and a BIOS ROM. However, Farcall is not a cycle-accurate emulator and currently lacks support for hard drives, EGA, or VGA modes. Its next goal is to support the protected mode of the 286 processor, followed by <strong>Windows 3.1</strong>.
        <p>Source : <a href="https://korben.info/farcall-emulateur-dos-ssh.html?utm_source=rss&amp;utm_medium=feed&amp;utm_campaign=flux-principal">Korben</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>RTX 5060 and RX 9060 XT Running on Windows 7 via Unofficial DirectX 12 Port</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/db04ff1ac4bc</link><guid isPermaLink="false">eink-db04ff1ac4bc-en</guid><pubDate>Fri, 20 Nov 2026 07:55:56 GMT</pubDate><category>maker</category><media:content url="https://korben.info/directx-12-windows-7-wddm2win7/directx-12-windows-7-wddm2win7-1.png" medium="image"/><description><![CDATA[Three developers have created an unofficial WDDM 2 port for Windows 7, enabling modern GPUs to run DirectX 12 applications.]]></description><content:encoded><![CDATA[<p><img src="https://korben.info/directx-12-windows-7-wddm2win7/directx-12-windows-7-wddm2win7-1.png" alt=""></p><p><strong>Three developers have created an unofficial WDDM 2 port for Windows 7, enabling modern GPUs to run DirectX 12 applications.</strong></p><p>Three developers (EAZY BLACK, VisualEhrmanntraut, and kryptik) have created an unofficial port of WDDM 2 for Windows 7, enabling modern GPUs like the GeForce RTX 5060, Radeon RX 9060 XT, or Intel UHD Graphics 770 to run DirectX 12 applications. This workaround allows outdated operating systems to recognize and utilize recent graphics cards, which would otherwise be incompatible.</p><p>WDDM (Windows Display Driver Model) is Microsoft’s framework for managing interactions between graphics drivers and the operating system. WDDM 2.0, introduced with Windows 10, includes support for Direct3D 12, a high-performance graphics API. While Microsoft released a user-mode DirectX 12 runtime for Windows 7 in 2019, it required individual applications to integrate it, limiting its practical use. This unofficial port restores full WDDM 2.0 functionality on Windows 7.</p><p>Tests confirm compatibility with RTX 40 and 50 series, Radeon RX 7000 and RX 9000 GPUs, as well as Intel UHD and Xe integrated graphics. Installation involves running a main package with trusted prompts, followed by adding the specific GPU driver via Device Manager. However, the project is experimental: no source code, license, or technical documentation is available on the GitHub repository.</p><p>The risks are significant—Windows 7 has been unsupported since January 2020, and its Extended Security Updates (ESU) ended in January 2023. The developers advise using this solution only on a dedicated test machine, disconnected from the network and without critical data. A full system backup is strongly recommended before proceeding.
        <p>Source : <a href="https://korben.info/directx-12-windows-7-wddm2win7.html?utm_source=rss&amp;utm_medium=feed&amp;utm_campaign=flux-principal">Korben</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>An Online Game That Crashes Real Microcontrollers</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/65de74796278</link><guid isPermaLink="false">eink-65de74796278-en</guid><pubDate>Fri, 20 Nov 2026 06:00:17 GMT</pubDate><category>maker</category><media:content url="https://hackaday.com/wp-content/uploads/2026/11/brett-e1795068033981.webp" medium="image"/><description><![CDATA[An online game where players crash microcontrollers to access virtual toilets, with a direct link to real hardware.]]></description><content:encoded><![CDATA[<p><img src="https://hackaday.com/wp-content/uploads/2026/11/brett-e1795068033981.webp" alt=""></p><p><strong>An online game where players crash microcontrollers to access virtual toilets, with a direct link to real hardware.</strong></p><p>Developer <strong>Ani</strong> has created <strong>Kacken Open Air (KOA)</strong>, an online game where players compete for access to virtual toilets, but with a twist: each in-game action directly impacts real microcontrollers.</p><p>The concept is straightforward: there are more players than available toilets. To claim one, participants must "knock" on virtual doors, which adds computational load to the linked microcontrollers (ESP8266). The more a player knocks, the slower the microcontroller becomes, eventually booting out the current occupant. Continuing the assault overloads the microcontroller, causing it to crash and reboot. In-game, this results in the virtual toilet collapsing.</p><p>The setup runs on a cluster of diverse microcontrollers: an ESP32-P4, multiple ESP8266s, Raspberry Pi Pico 2s, Arduino Due boards, and an STM8 acting as a hardware watchdog to keep the system stable. The game is playable directly in a web browser, with an interface in German and an automatic translation option via Google Translate.
        <p>Source : <a href="https://hackaday.com/2026/11/19/crashing-microcontrollers-competitively/">Hackaday</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>DabbleCAD: An Offline, Open-Source CAD Tool for 3D Design and Laser Cutting</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/39275de62d90</link><guid isPermaLink="false">eink-39275de62d90-en</guid><pubDate>Fri, 20 Nov 2026 03:00:57 GMT</pubDate><category>maker</category><media:content url="https://hackaday.com/wp-content/uploads/2026/11/screenshot-3d-design-e1795089403514.png" medium="image"/><description><![CDATA[A lightweight, offline, open-source CAD tool supporting 3D design and laser cutting workflows.]]></description><content:encoded><![CDATA[<p><img src="https://hackaday.com/wp-content/uploads/2026/11/screenshot-3d-design-e1795089403514.png" alt=""></p><p><strong>A lightweight, offline, open-source CAD tool supporting 3D design and laser cutting workflows.</strong></p><p>DabbleCAD breaks away from the trend of web-based and browser-dependent CAD tools by offering a simple, offline alternative. This open-source software enables 3D modeling through parametric design, much like TinkerCAD, with standard tools for stretching, deforming, and punching holes. It supports exporting files in STL or 3MF formats, ready for slicers in 3D printing workflows.</p><p>Additionally, DabbleCAD includes features for laser cutting. Users can import artwork or draw directly, then export G-code or send commands directly to a GRBL-compatible laser cutter via USB. A relief engraving option is also available: 3D models or images can be converted into height maps for relief engraving, provided the necessary tools are available.</p><p>While it doesn’t introduce groundbreaking new methods, DabbleCAD provides a practical and portable solution, free from the hassles of internet connectivity, user accounts, or cloud services. Its approach is designed to be accessible to makers and hobbyists looking to quickly design parts or prototypes without relying on online tools.
        <p>Source : <a href="https://hackaday.com/2026/11/19/dabblecad-takes-design-offline/">Hackaday</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Smart Garage Opener: Bypassing Remote Servers to Stop Constant Reboots</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/846dadb75d31</link><guid isPermaLink="false">eink-846dadb75d31-en</guid><pubDate>Fri, 20 Nov 2026 00:00:48 GMT</pubDate><category>maker</category><media:content url="https://hackaday.com/wp-content/uploads/2026/11/f2712a2c7f85e0ec8940.png" medium="image"/><description><![CDATA[A user bypasses mandatory connections to remote servers on a smart garage door opener by extracting its MQTT key and simulating a local server.]]></description><content:encoded><![CDATA[<p><img src="https://hackaday.com/wp-content/uploads/2026/11/f2712a2c7f85e0ec8940.png" alt=""></p><p><strong>A user bypasses mandatory connections to remote servers on a smart garage door opener by extracting its MQTT key and simulating a local server.</strong></p><p>Some smart garage door openers, like the <strong>LiftMaster MyQ Home Bridge</strong>, enforce mandatory connections to remote servers via encrypted MQTT. When these connections fail, the device reboots repeatedly to restore contact, a frustrating behavior when offline or restricted to an IoT VLAN.</p><p><strong>Kusha</strong> solved this by extracting the MQTT encryption key stored on the device’s <strong>Cortex-M4</strong> microcontroller. Using the accessible <strong>SWD (Serial Wire Debug)</strong> interface on the PCB, he read the required data with custom wiring and Arduino code. After retrieving the key, he set up a <strong>local MQTT broker</strong> (such as Mosquitto) with the key installed to simulate a valid server. A Docker configuration file is provided for easy deployment.</p><p>To redirect connection attempts to this local server, Kusha used a <strong>local DNS server</strong> to map the manufacturer’s domains (<strong>mqdevice.com</strong>) to the local broker’s address. The garage door opener now believes it is always online and stops rebooting, even on an isolated IoT network.
        <p>Source : <a href="https://hackaday.com/2026/11/19/preventing-a-garage-door-opener-from-rebooting-when-offline/">Hackaday</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Inflatable Soft Robot 100x Faster Using Steam Engine Principles</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/a2603d43e121</link><guid isPermaLink="false">eink-a2603d43e121-en</guid><pubDate>Thu, 19 Nov 2026 23:30:00 GMT</pubDate><category>maker</category><media:content url="https://i0.wp.com/makezine.com/wp-content/uploads/2026/11/Jacobs-Soft-Robot.jpg?fit=750%2C393&amp;ssl=1" medium="image"/><description><![CDATA[UC San Diego engineers unveiled a six-legged inflatable robot 100 times faster than its predecessor, borrowing steam engine mechanics to overcome pressure challenges.]]></description><content:encoded><![CDATA[<p><img src="https://i0.wp.com/makezine.com/wp-content/uploads/2026/11/Jacobs-Soft-Robot.jpg?fit=750%2C393&amp;ssl=1" alt=""></p><p><strong>UC San Diego engineers unveiled a six-legged inflatable robot 100 times faster than its predecessor, borrowing steam engine mechanics to overcome pressure challenges.</strong></p><p>Engineers at the University of California San Diego (UC San Diego Jacobs School of Engineering) have unveiled an upgraded version of their six-legged inflatable soft robot, first introduced in 2022. The new model, which lacks complex electronics, is 100 times faster than its predecessor. To achieve this speed, the team increased the internal pressure of the robot, requiring a redesign of its structure.</p><p>Rather than relying on advanced electronics, the researchers drew inspiration from steam engine mechanics. This approach allowed them to address the challenges posed by higher pressure while maintaining the lightweight and flexible nature of soft robots. Their goal is to create a machine capable of operating in environments where traditional, rigid, and electronic robots struggle.</p><p>The project highlights the potential of soft robots, particularly their ability to move quickly despite their flexible structure. The team has not yet disclosed specific applications or next steps for the robot’s development.</p>
        <p>Source : <a href="https://makezine.com/article/technology/robotics/soft-robot-is-faster-stronger-than-before/">Make:</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Restoring a BBC Micro with Sideways ROM/RAM Expansion</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/fc1892051e52</link><guid isPermaLink="false">eink-fc1892051e52-en</guid><pubDate>Thu, 19 Nov 2026 22:30:06 GMT</pubDate><category>maker</category><media:content url="https://cdn-blog.adafruit.com/uploads/2026/11/restoring-a-bbc-microcomputer-part-4-going-sideways.jpg" medium="image"/><description><![CDATA[An enthusiast upgrades a BBC Micro Model B with a sideways expansion board to add multiple ROMs and RAM.]]></description><content:encoded><![CDATA[<p><img src="https://cdn-blog.adafruit.com/uploads/2026/11/restoring-a-bbc-microcomputer-part-4-going-sideways.jpg" alt=""></p><p><strong>An enthusiast upgrades a BBC Micro Model B with a sideways expansion board to add multiple ROMs and RAM.</strong></p><p>The BBC Micro Model B, a classic British computer from the 1980s, is being restored by Andrew Dorey with the addition of a sideways expansion board. This upgrade bypasses the hardware limitation of the 6502 processor, which can only address 64 KB of memory at once. Acorn reserved a 16 KB window (from &amp;8000 to &amp;BFFF) to share up to sixteen 16 KB ROMs. A latch at address &amp;FE30 selects the active ROM, while the MOS reads each ROM’s header at power-up to build an installation table.</p> <p>The expansion board, a <strong>Watford Electronics 12 ROM card</strong>, provides twelve ROM slots or eleven ROMs paired with two 8 KB SRAM chips, totaling 16 KB of sideways RAM. Installation required cleaning, reflowing solder joints, replacing an S21 cable, repairing 5 V and ground connections, and adding DIP sockets for an extra 5 mm of clearance. Several ROMs (assembler, monitor, word processor) were burned onto 27C128 and 27C256 EPROMs, though many second-hand chips proved faulty or mislabeled.</p> <p>An external three-cell NiMH battery keeps the RAM alive, and a 470 μF capacitor fixed a 120 μs voltage drop at power-off that was corrupting bytes. The project is documented at andrewdorey.com.
        <p>Source : <a href="https://blog.adafruit.com/2026/11/19/restoring-a-bbc-microcomputer-going-sideways/">Adafruit</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Why the 2014 Nobel Prize Rewarded the Blue LED</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/6b00bfb6e127</link><guid isPermaLink="false">eink-6b00bfb6e127-en</guid><pubDate>Thu, 19 Nov 2026 22:24:30 GMT</pubDate><category>maker</category><media:content url="https://cdn-blog.adafruit.com/uploads/2026/11/why-did-only-blue-led-win-the-nobel-prize.jpg" medium="image"/><description><![CDATA[An open-source project breaks down the physics of blue LEDs and offers two hands-on breadboard demos.]]></description><content:encoded><![CDATA[<p><img src="https://cdn-blog.adafruit.com/uploads/2026/11/why-did-only-blue-led-win-the-nobel-prize.jpg" alt=""></p><p><strong>An open-source project breaks down the physics of blue LEDs and offers two hands-on breadboard demos.</strong></p><p>The 2014 Nobel Prize in Physics recognized the invention of the blue LED, a breakthrough after decades of research. Unlike red or green LEDs, blue LEDs require a wide-bandgap semiconductor material capable of emitting high-energy photons. <strong>Gallium nitride (GaN)</strong>, the material used, was notoriously difficult to grow without defects. In 1986, Isamu Akasaki and Hiroshi Amano successfully cultivated high-quality GaN crystals on sapphire substrates, while Shuji Nakamura later improved their efficiency by activating p-type doping through thermal annealing, making commercial blue LEDs viable.</p><h2>Two hands-on demos</h2><p>The project includes two simple experiments to illustrate these principles. The first involves measuring the forward voltage across red, green, and blue LEDs with a multimeter, highlighting the bandgap differences: about 1.8 V for red, 2 V for green, and 3 V for blue. The second demo uses a <strong>YAG:Ce phosphor film</strong> (cerium-doped yttrium-aluminum garnet) placed over a blue LED to convert part of the light into complementary wavelengths, producing white light—the same principle behind modern LED bulbs.</p>
        <p>Source : <a href="https://blog.adafruit.com/2026/11/19/why-did-only-blue-led-win-the-nobel-prize/">Adafruit</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>AI Speeds Up Reverse Engineering with Open Source Tools</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/2dfa47bf8376</link><guid isPermaLink="false">eink-2dfa47bf8376-en</guid><pubDate>Thu, 19 Nov 2026 22:17:09 GMT</pubDate><category>maker</category><media:content url="https://cdn-blog.adafruit.com/uploads/2026/11/llm-assisted-reverse-engineering.jpg" medium="image"/><description><![CDATA[A developer automates binary reverse engineering with an AI agent, cutting weeks of manual work down to hours.]]></description><content:encoded><![CDATA[<p><img src="https://cdn-blog.adafruit.com/uploads/2026/11/llm-assisted-reverse-engineering.jpg" alt=""></p><p><strong>A developer automates binary reverse engineering with an AI agent, cutting weeks of manual work down to hours.</strong></p><p>Reverse engineering embedded software, often a tedious process, is being transformed by combining an AI agent with open source tools. <strong>ioma8</strong>, with two decades of experience in hardware and software dissection, explains on <em>slowchronicle.com</em> how their workflow now includes an AI coding assistant. This assistant directly analyzes binaries, guided by specific skills, and generates a code overview far faster than manual methods.</p><p>Two key modules are shared as open source. The first details reverse engineering tools like <strong>radare2</strong> and an MCP bridge to <strong>Ghidra</strong>, along with 22 Ghidra scripts for targeted disassembly. Without these scripts, the agent often quits early because it doesn’t know how to run the disassembly step. The second module, called <strong>hypothesis-debugging</strong>, forces the AI to state one testable hypothesis at a time, logged in a CSV loop. This method, inspired by the scientific approach, prevents random guesses and endless loops. Findings are written to markdown files in real time.</p><p>Technical details and links to resources are available on <em>slowchronicle.com</em> and the associated GitHub repository.
        <p>Source : <a href="https://blog.adafruit.com/2026/11/19/llm-assisted-reverse-engineering/">Adafruit</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Automating Appliances Without Timers Using Home Assistant</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/5744bfcea503</link><guid isPermaLink="false">eink-5744bfcea503-en</guid><pubDate>Thu, 19 Nov 2026 22:00:21 GMT</pubDate><category>Dashboards</category><media:content url="https://static0.xdaimages.com/wordpress/wp-content/uploads/wm/2026/11/tapo-smart-plug-energy-monitoring-7.jpeg" medium="image"/><description><![CDATA[Replace fixed timers with power draw sensors for more reliable home automations.]]></description><content:encoded><![CDATA[<p><img src="https://static0.xdaimages.com/wordpress/wp-content/uploads/wm/2026/11/tapo-smart-plug-energy-monitoring-7.jpeg" alt=""></p><p><strong>Replace fixed timers with power draw sensors for more reliable home automations.</strong></p><p>Traditional Home Assistant automations often rely on static timers, such as assuming a washing machine cycle lasts exactly 45 minutes. This approach fails when eco-modes, partial loads, or unexpected pauses occur. A better solution is to use smart plugs that monitor real-time power consumption, enabling precise detection of appliance states.</p> <h2>Washing and drying: tracking cycles by power draw</h2> <p>Washing machines and dryers have variable consumption profiles, with heating, washing, and pause phases. A plug like the <strong>Tapo P110</strong> (which monitors energy) can trigger an automation as soon as consumption exceeds 5 W, indicating the start of a cycle. To avoid false detections during brief pauses (where consumption drops to 0 W), a delay condition is added: consumption must remain below 2 W for at least 5 minutes before sending a notification or activating a smart speaker to announce cycle completion.</p> <p>This method avoids errors from rigid timers and adapts to varying cycle durations. A persistent notification can also appear in the Home Assistant interface.</p> <h2>Coffee machines: preventing overheating and waste</h2> <p>Manual or drip coffee machines often lack automatic shut-off. By analyzing power draw, Home Assistant distinguishes a heating phase (over 800 W), a keep-warm state (around 50 W), and inactivity (2 W). The automation cuts power as soon as consumption stays below 2 W for 15 minutes, reducing fire risks and idle energy use. This logic also applies to espresso machines.</p> <p>Dehumidifiers experience a sharp drop in consumption from 200 W to 2 W when their tank is full. Home Assistant waits for this drop to last over 30 minutes before sending an alert to empty the reservoir, preventing overflows.</p> <p>This approach turns non-smart appliances into intelligent nodes without relying on arbitrary timers. It improves the reliability of notifications, safety cutoffs, and energy tracking. However, it’s important to check the maximum power rating of the smart plug, as high-draw appliances like dryers or kettles may approach the limits of standard models.</p>
        <p>Source : <a href="https://www.xda-developers.com/home-assistant-automations-power-draw-instead-timers-control-appliances/">XDA</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>A Giant LED Matrix Cube Powered by Raspberry Pi CM4</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/3df7be2b27ae</link><guid isPermaLink="false">eink-3df7be2b27ae-en</guid><pubDate>Thu, 19 Nov 2026 21:00:39 GMT</pubDate><category>maker</category><media:content url="https://hackaday.com/wp-content/uploads/2026/11/9683211794920017950.jpg" medium="image"/><description><![CDATA[A 6-sided LED cube featuring 24,576 LEDs, driven by a Raspberry Pi Compute Module 4, displaying animations, games, and music visualizations.]]></description><content:encoded><![CDATA[<p><img src="https://hackaday.com/wp-content/uploads/2026/11/9683211794920017950.jpg" alt=""></p><p><strong>A 6-sided LED cube featuring 24,576 LEDs, driven by a Raspberry Pi Compute Module 4, displaying animations, games, and music visualizations.</strong></p><p>This impressive LED cube, designed by <strong>Ross Harvey</strong>, features six faces entirely covered with HUB75 LED panels. Each face contains 64 x 64 LEDs spaced 2.5 mm apart, forming a 160 mm square. The cube totals <strong>24,576 individually addressable LEDs</strong>, providing a massive and precise display surface.</p> <p>To soften the LED glare and create a more uniform brightness, Harvey added a cell-cast acrylic diffuser sheet to each face. Inside the cube, a <strong>Raspberry Pi Compute Module 4</strong> handles the control, paired with a custom HUB75 driver board. A power and boost board, powered by a 2-cell lithium-ion battery, supplies the necessary energy. Two speakers and a microphone are also included, enabling audio interactions and sound reactivity.</p> <p>The cube supports a wide range of content: graphic animations, video games, music visualizers, and even an elegant world clock. Harvey also offers a ready-to-use version for those who prefer to skip the full build. For fans of more complex geometric shapes, <strong>Greg Davill</strong> created a glowing D20 (icosahedral die), a well-known project in the community.
        <p>Source : <a href="https://hackaday.com/2026/11/19/a-big-ol-led-matrix-cube/">Hackaday</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>3D-Printed Case for Adafruit 64×32 LED Matrix &amp; MatrixPortal M4</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/cfd4947fd321</link><guid isPermaLink="false">eink-cfd4947fd321-en</guid><pubDate>Thu, 19 Nov 2026 20:00:43 GMT</pubDate><category>DIY &amp; projects</category><media:content url="https://cdn-blog.adafruit.com/uploads/2026/11/2024-11-01_54cc1d882d01c.jpg-300x218.webp" medium="image"/><description><![CDATA[An open-source 3D-printed case to mount a 64×32 LED matrix and MatrixPortal M4 board, perfect for custom dashboards.]]></description><content:encoded><![CDATA[<p><img src="https://cdn-blog.adafruit.com/uploads/2026/11/2024-11-01_54cc1d882d01c.jpg-300x218.webp" alt=""></p><p><strong>An open-source 3D-printed case to mount a 64×32 LED matrix and MatrixPortal M4 board, perfect for custom dashboards.</strong></p><p>This 3D-printed case is designed to hold a P4 64×32 RGB LED matrix panel (256×128mm), an Adafruit MatrixPortal M4 board at the back, and an Adafruit diffuser (product 4749, 260×130mm) at the front. The enclosure is split into two parts to fit smaller printers like the Prusa Mini. A remixed version adds support for mid-point connectors, while screws and an acrylic diffuser hold-down (using double-sided adhesive pads) secure the assembly.</p><p>Recommended print settings include a 0.16mm layer height, two walls, 15% infill, and an estimated print time of ~6.2 hours for two plates, using ~188g of PLA. This case is part of the Adafruit ecosystem for dashboard projects, combining LED displays with microcontrollers for custom displays, such as a metro or bus schedule board.</p><p>The STL file is available as open source and has received a 5.0★ rating from 5 reviews. It is a community-driven project, remixed by Cpt.Omarica from the original design by tip2tail.</p>
        <p>Source : <a href="https://blog.adafruit.com/2026/11/19/case-for-adafruit-64x32-led-matrix-matrixportal-m4-diffuser-housing-3dthursday-3dprinting/">Adafruit</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>JieLi AC707N: A $10 Digital Badge Turned Open-Source Hacking Platform</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/d4171a4083cb</link><guid isPermaLink="false">eink-d4171a4083cb-en</guid><pubDate>Thu, 19 Nov 2026 18:00:00 GMT</pubDate><category>DIY &amp; projects</category><description><![CDATA[A $10 digital badge is transformed into an open-source hacking platform through reverse engineering.]]></description><content:encoded><![CDATA[<p><strong>A $10 digital badge is transformed into an open-source hacking platform through reverse engineering.</strong></p><p>The <strong>JieLi AC707N</strong>, a digital badge priced around $10, packs surprisingly capable hardware. Part of the <strong>BR35 family</strong>, it features a 360×360 pixel circular display driven over <strong>QSPI</strong>, a capacitive touch controller <strong>AXS5106L</strong>, two physical buttons, 4 MB of flash memory, and <strong>Bluetooth Low Energy</strong> connectivity.</p><p>A developer known as <strong>[Sabas]</strong> reverse-engineered the device by first physically disassembling it, then decompiling the accompanying Android app using <strong>Ghidra</strong>. The analysis uncovered a lack of authentication for firmware downloads, enabling custom firmware to be flashed onto sealed, factory-fresh units. An open-source <strong>SDK</strong> was subsequently developed to streamline custom firmware creation.</p><p>A preliminary alternative firmware is already available on <strong>GitHub</strong>. It repurposes the badge as a Bluetooth mouse or scanner, but the potential extends far beyond. Users can now modify the code for their own applications, such as smart home interfaces, interactive badges, or diagnostic tools.
        <p>Source : <a href="https://hackaday.com/2026/11/19/digital-badge-gets-reverse-engineered/">Hackaday</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>A Solar-Powered Robotic Jellyfish Built Over Three Years of Tinkering</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/539118461eb0</link><guid isPermaLink="false">eink-539118461eb0-en</guid><pubDate>Thu, 19 Nov 2026 15:10:31 GMT</pubDate><category>maker</category><description><![CDATA[A hobbyist spent three years developing a solar-powered robotic jellyfish, using gears and motors to mimic natural swimming, after numerous prototypes.]]></description><content:encoded><![CDATA[<p><strong>A hobbyist spent three years developing a solar-powered robotic jellyfish, using gears and motors to mimic natural swimming, after numerous prototypes.</strong></p><p>A maker drew inspiration from the jellyfish’s lack of brain, heart, or bones to create a robotic version that swims autonomously in an aquarium, powered solely by solar energy. After three years of trial and error, the project culminated in a functional prototype using gears, motors, and custom electronics.</p><p>The journey began with a simpler creation: Rocky, a mechanical "pet rock" featuring a solar panel, rechargeable battery, and motor to shuffle around on legs. This early attempt provided foundational robotics experience before tackling the jellyfish design. The final robot relies on an <strong>ATtiny85 microcontroller</strong> to manage energy harvesting from tiny solar panels. Energy is slowly accumulated in a capacitor and released in short bursts to power the motor, overcoming the panels’ limited current output.</p><p>Propulsion mimics real jellyfish swimming through a unique mechanism. A saw-shaped cam pushes hinged rods outward, stretching a flexible silicone bell. When released, the bell springs back, pushing water downward to generate thrust. The system includes a 7:1 gear reduction, carbon fiber rods, and a custom-molded silicone exterior. A laser-cut acrylic plate separates the wet propulsion components from the electronics, while gaskets and a rotary shaft seal prevent water ingress.</p><p>Buoyancy adjustments were critical; the lightweight upper dome was replaced with a thicker resin version to add weight without compromising the jellyfish’s appearance. The finished robot features five solar panels and swims under bright artificial lighting, though this setup isn’t energy-efficient. The creator emphasized the goal was to bring a long-held idea to life, not to optimize power usage, despite years of failed prototypes.
        <p>Source : <a href="https://www.hackster.io/news/a-jellyfish-made-of-motors-gears-and-solar-panels-44fb901299b1">Hackster</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>KUKA Robots Carve Marble Sculptures with Precision</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/70a2f48a7c9b</link><guid isPermaLink="false">eink-70a2f48a7c9b-en</guid><pubDate>Thu, 19 Nov 2026 15:05:20 GMT</pubDate><category>maker</category><media:content url="https://cdn-blog.adafruit.com/uploads/2026/11/we-taught-robots-to-carve-marble-sculptures.jpg" medium="image"/><description><![CDATA[Industrial KUKA KR240 robots autonomously carve marble blocks into finished sculptures without human oversight.]]></description><content:encoded><![CDATA[<p><img src="https://cdn-blog.adafruit.com/uploads/2026/11/we-taught-robots-to-carve-marble-sculptures.jpg" alt=""></p><p><strong>Industrial KUKA KR240 robots autonomously carve marble blocks into finished sculptures without human oversight.</strong></p><p><strong>Monumental Labs</strong>, based in Brooklyn, employs KUKA KR240 robots to autonomously carve marble blocks into detailed sculptures. Demonstrated at <strong>Embodied AI NYC</strong>, the system processes a 3D model and a raw stone block to produce a finished piece, as shown in a trick where a Stanford bunny emerges from a top hat carved out of marble.</p><p>The process starts with a <strong>sparse voxel grid</strong> representing the initial block. The software then computes a <strong>signed distance field (SDF)</strong> with normals and gradients to guide the carving. Roughing passes remove bulk material in isoplanar layers, supplemented by tilted passes to reach undercuts like chins or ears. Finishing refines the sculpture in offset layers, with the final layer revealing the intricate design.</p><p>Toolpaths are optimized using <strong>geodesic distances</strong> (measured along the surface rather than straight-line), and the system simulates all movements to ensure collision-free orientations for the robot, tool, and the marble block—whose machining cost can reach one million dollars. Further details are available at <strong>eleanorwiesler.com</strong>.</p>
        <p>Source : <a href="https://blog.adafruit.com/2026/11/19/teaching-robots-to-carve-marble-sculptures/">Adafruit</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Open-Source I2C TPM 2.0 Security Module for Raspberry Pi Launches Crowdfunding</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/94a251fb0aaf</link><guid isPermaLink="false">eink-94a251fb0aaf-en</guid><pubDate>Thu, 19 Nov 2026 14:49:34 GMT</pubDate><category>maker</category><media:content url="https://cdn-blog.adafruit.com/uploads/2026/11/raspberry-pi-sbcs-get-an-open-source-hardware-i2c-tpm-2-0-se.jpg" medium="image"/><description><![CDATA[An open-source security module for Raspberry Pi, featuring a TPM 2.0 chip, is now available for pre-order with delivery expected in March 2027.]]></description><content:encoded><![CDATA[<p><img src="https://cdn-blog.adafruit.com/uploads/2026/11/raspberry-pi-sbcs-get-an-open-source-hardware-i2c-tpm-2-0-se.jpg" alt=""></p><p><strong>An open-source security module for Raspberry Pi, featuring a TPM 2.0 chip, is now available for pre-order with delivery expected in March 2027.</strong></p><p>A new open-source security module for Raspberry Pi single board computers has been announced by ANAVI. Called the <strong>ANAVI I2C TPM 2.0</strong>, it features the <strong>Infineon Optiga SLB 9673</strong> chip, designed for secure cryptographic key storage. Unlike software-based solutions, this component keeps keys in hardware, preventing access from the SD card or operating system.</p><p>The module connects via the <strong>I2C</strong> bus, a standard interface for electronic device communication. It follows an earlier version that used an <strong>SPI</strong> connection with the SLB 9672 chip. Its applications include disk encryption, secure key storage, and hardware-based random number generation, which are critical for cryptographic operations.</p><p>The Linux kernel in compatible Raspberry Pi distributions (such as Raspberry Pi OS) already includes a <strong>device tree overlay</strong> for the SLB 9673, ensuring immediate compatibility without additional software installation. A detailed tutorial, published by Leon ANAVI, explains how to set up an encrypted data partition using the module. The KiCad design files are available on GitHub under the <strong>CC BY-SA 4.0</strong> license.</p><p>The module is offered in two versions: a horizontal one for $20 or a vertical one for $21. Pre-orders are open on <strong>Crowd Supply</strong>, with deliveries scheduled to begin on March 18, 2027.
        <p>Source : <a href="https://blog.adafruit.com/2026/11/19/raspberry-pi-sbcs-get-an-open-source-hardware-i2c-tpm-2-0-security-module-crowdfunding/">Adafruit</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Python on RISC-V: 1,500 Packages Ready to Use</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/fb2b03e3a68e</link><guid isPermaLink="false">eink-fb2b03e3a68e-en</guid><pubDate>Thu, 19 Nov 2026 14:46:13 GMT</pubDate><category>maker</category><media:content url="https://cdn-blog.adafruit.com/uploads/2026/11/python-on-risc-v-1500-packages-and-counting.jpg" medium="image"/><description><![CDATA[Over 1,500 native Python packages are now available for RISC-V, speeding up installation on this emerging architecture.]]></description><content:encoded><![CDATA[<p><img src="https://cdn-blog.adafruit.com/uploads/2026/11/python-on-risc-v-1500-packages-and-counting.jpg" alt=""></p><p><strong>Over 1,500 native Python packages are now available for RISC-V, speeding up installation on this emerging architecture.</strong></p><p>The RISE project reports that over 1,500 Python packages with native code (C, C++, Rust, or Fortran) are now available as <strong>wheels</strong> for the RISC-V architecture. These precompiled files eliminate the need for lengthy source-based compilation. On a <strong>VisionFive 2</strong>, installing a package like NumPy drops from 15 minutes to just 25 seconds thanks to this optimization.</p><p>Major scientific libraries (<strong>NumPy, SciPy, pandas</strong>), encryption tools (<strong>cryptography</strong>), database engines (<strong>PyArrow</strong>), and AI frameworks (<strong>PyTorch, vLLM</strong>) are now among the supported packages. The RISC-V wheels for PyTorch 2.13.0 were built natively and validated through 212,038 test cases.</p><p>Contributions follow automated <strong>GitHub Actions workflows</strong>. Each package is compiled on RISC-V hardware, tested in an isolated environment, and published with versioned patches. Despite progress, 519 packages still lack functional wheels. The long-term goal is for upstream projects to directly publish their own RISC-V wheels, making RISE’s intermediate builds unnecessary.</p>
        <p>Source : <a href="https://blog.adafruit.com/2026/11/19/python-on-risc-v-1500-packages-and-counting/">Adafruit</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>A Paperclip-Weight Flying Robot the Size of a Microcassette</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/0c72de0e76ef</link><guid isPermaLink="false">eink-0c72de0e76ef-en</guid><pubDate>Thu, 19 Nov 2026 14:35:44 GMT</pubDate><category>maker</category><description><![CDATA[MIT researchers have built a sub-gram flying robot that navigates autonomously using onboard sensors.]]></description><content:encoded><![CDATA[<p><strong>MIT researchers have built a sub-gram flying robot that navigates autonomously using onboard sensors.</strong></p><p>Researchers at MIT have developed a sub-gram flying robot inspired by insects, measuring roughly the size of a microcassette. Its four wings are powered by soft artificial muscles that contract and expand rapidly, enabling stable flight and precise maneuvers.</p><h2>Onboard navigation and autonomy</h2><p>Previous versions of these robots relied on external motion-capture cameras for positioning, restricting their use to controlled environments. The team overcame this by integrating a custom electronics package weighing just 244 milligrams, including a microcontroller and three sensors: an inertial measurement unit for orientation, an optical flow sensor for horizontal movement, and a time-of-flight distance sensor for altitude.</p><p>To mitigate measurement errors that could cause drift, the researchers designed a lightweight algorithm that continuously estimates the robot’s position and orientation. Instead of complex calculations, the system performs simpler, more frequent operations better suited to the tiny processor’s capabilities.</p><p>In flight tests, the robot successfully hovered at various heights and followed predefined paths with positional errors of just a few centimeters. It also demonstrated obstacle avoidance under joystick control, navigating around plants before landing on a sunflower.</p><h2>Current limitations and future work</h2><p>Despite these breakthroughs, the robot still requires a wired power supply. Future developments aim to add a battery and a camera for object recognition and autonomous obstacle avoidance. These upgrades are crucial before the robot can perform tasks like artificial pollination or machinery inspection in real-world conditions.
        <p>Source : <a href="https://www.hackster.io/news/this-flying-robot-weighs-less-than-a-paperclip-a64d21f4bb8a">Hackster</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Temperature and Humidity Sensors Comparison for ESP32 and RP2040</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/da4e7966053c</link><guid isPermaLink="false">eink-da4e7966053c-en</guid><pubDate>Thu, 19 Nov 2026 14:23:28 GMT</pubDate><category>maker</category><media:content url="https://randomnerdtutorials.com/wp-content/uploads/2026/11/ESP32-Temperature-Sensors-Comparison.jpg" medium="image"/><description><![CDATA[A comprehensive guide to selecting temperature and humidity sensors for microcontroller projects like ESP32 or RP2040.]]></description><content:encoded><![CDATA[<p><img src="https://randomnerdtutorials.com/wp-content/uploads/2026/11/ESP32-Temperature-Sensors-Comparison.jpg" alt=""></p><p><strong>A comprehensive guide to selecting temperature and humidity sensors for microcontroller projects like ESP32 or RP2040.</strong></p><p>Temperature and humidity sensors are essential for many electronics projects, especially when using microcontrollers like the <strong>ESP32</strong>, <strong>Raspberry Pi Pico</strong>, or <strong>Arduino</strong>. This comparison evaluates the most popular models based on accuracy, resolution, communication protocol, temperature range, ease of use, and library support.</p> <h2>Available sensor families</h2> <p>Several sensor families are covered, including the <strong>AHT</strong> series (AHT10, AHT20, AHT21), <strong>DHT</strong> series (DHT11, DHT22, AM2320), and <strong>SHT</strong> series (SHT30 to SHT45). The <strong>BME</strong> sensors (BME280, BME680, BME688) also measure atmospheric pressure, while <strong>SCD</strong> models (SCD40, SCD41) include CO₂ detection. Other options include thermocouples, non-contact infrared sensors like the <strong>MLX90614</strong>, and analog sensors.</p> <p>The choice depends on the application: a sensor like the <strong>BME280</strong>, which is I2C-compatible and easy to use, is ideal for learning IoT concepts. For projects requiring high precision, <strong>Sensirion SHT4x</strong> sensors (notably the SHT41) offer an excellent accuracy-to-price ratio, though they do not measure pressure.</p> <h2>Selection criteria and resources</h2> <p>The guide provides comparison tables including supply voltage, temperature range, accuracy, and typical price. Support for <strong>Arduino/PlatformIO</strong>, <strong>MicroPython</strong>, and <strong>CircuitPython</strong> libraries is also assessed, a key factor for easier development and debugging. For example, the <strong>DHT22</strong> is often preferred for its low cost, while the <strong>SHT41</strong> is recommended for applications requiring high reliability.</p> <p>Additional resources, such as detailed tutorials for the <strong>ESP32</strong>, <strong>Raspberry Pi Pico</strong>, or <strong>Arduino</strong>, are available to help with sensor selection and implementation.</p>
        <p>Source : <a href="https://randomnerdtutorials.com/comparison-guide-temperature-humidity-sensors/">Random Nerd Tutorials</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Waveshare Unveils ESP32-S31 Touchscreen Devkit with 4.3&quot; Display and Optional DIN Rail Case</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/704436ca4366</link><guid isPermaLink="false">eink-704436ca4366-en</guid><pubDate>Thu, 19 Nov 2026 14:19:12 GMT</pubDate><category>displays</category><media:content url="https://www.cnx-software.com/wp-content/uploads/2026/11/Wavesahre-ESP32-S31-TOUCH-LCD-4.3C-720x485.jpg" medium="image"/><description><![CDATA[Waveshare releases an ESP32-S31-based development kit featuring a 4.3-inch touchscreen and RS485/CAN interfaces, available in three configurations.]]></description><content:encoded><![CDATA[<p><img src="https://www.cnx-software.com/wp-content/uploads/2026/11/Wavesahre-ESP32-S31-TOUCH-LCD-4.3C-720x485.jpg" alt=""></p><p><strong>Waveshare releases an ESP32-S31-based development kit featuring a 4.3-inch touchscreen and RS485/CAN interfaces, available in three configurations.</strong></p><p>Waveshare has launched the <strong>ESP32-S31-Touch-LCD-4.3C</strong>, a development kit powered by an ESP32-S31 module with 16MB flash and 16MB PSRAM. The kit features a 4.3-inch capacitive touchscreen LCD with 800×480 resolution and 65K colors, driven by an ST7262 controller over a 16-bit RGB565 interface. The IPS display offers a 160° viewing angle.</p><p>Audio capabilities include an Everest ES8389 codec, two microphones, and an 8 Ω, 2 W speaker. Connectivity options cover Wi-Fi 6 (802.11ax), Bluetooth 5.4 (LE Audio, direction finding, Mesh 1.1), and an 802.15.4 radio for Zigbee, Thread, and Matter. RS485 and CAN FD interfaces are available via terminal blocks, alongside six GPIOs (GPIO48–53). A USB Type-C port handles 5V power, programming, and serial debugging via a CH343P chip.</p><p>The board supports flexible power inputs: 7–36V via terminal block, 5V via USB Type-C, or an optional 3.7V lithium battery. An optional DIN rail enclosure and lithium-ion battery support are also provided. Waveshare offers sixteen low-level code examples for ESP-IDF 6.1, including firmware, schematics, and mechanical files on GitHub.</p><p>Three SKUs are available: standard without enclosure, with enclosure, or with enclosure and battery. Pricing starts at $45.99 on Waveshare’s store, $53.99 on AliExpress, and $71.99 on Amazon (enclosure-included version).
        <p>Source : <a href="https://www.cnx-software.com/2026/11/19/waveshare-esp32-s31-touch-lcd-4-3c-an-esp32-s31-4-3-inch-display-devkit-with-optional-din-rail-case/">CNX Software</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Favor Brunner: The Art of Mechatronics with Raspberry Pi</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/87a71909bb37</link><guid isPermaLink="false">eink-87a71909bb37-en</guid><pubDate>Thu, 19 Nov 2026 13:57:19 GMT</pubDate><category>maker</category><media:content url="https://www.raspberrypi.com/app/uploads/2026/11/IMG_7645-Large.jpeg" medium="image"/><description><![CDATA[A mechatronics engineer turns technical projects into artistic works using Raspberry Pi and an early passion for building.]]></description><content:encoded><![CDATA[<p><img src="https://www.raspberrypi.com/app/uploads/2026/11/IMG_7645-Large.jpeg" alt=""></p><p><strong>A mechatronics engineer turns technical projects into artistic works using Raspberry Pi and an early passion for building.</strong></p><p>Favor Brunner, a mechatronics engineer, views building as a form of artistic expression. His journey began at age seven with Lego, evolving into interactive projects thanks to <strong>Raspberry Pi</strong> boards gifted by his software engineer uncle. A primary school robotics program, funded through a barter of his mother’s original artwork, further nurtured his passion.</p><p>He first encountered Raspberry Pi in third grade via a Kano kit, assembled by his uncle. This introduction to a programmable computer sparked a sense of agency—allowing him to understand, modify, and create, while sharing memories with his cousins. His interest in robotics led to a YouTube and TikTok channel documenting his builds.</p><p>Among his standout projects is <strong>Run It Back NYC</strong>, a transcriber for New York City subway announcements using a <strong>Raspberry Pi</strong> and a touchscreen. The goal: translate and audibly relay incomprehensible MTA announcements for riders. Another project, a local <strong>AI chatbot named Baymax</strong>, served as a Mandarin tutor, leveraging on-device artificial intelligence.</p><p>These creations reflect his approach: addressing real-world problems through accessible solutions. His work is shared on <strong>TikTok and YouTube</strong>, where he continues to publish new builds.</p>
        <p>Source : <a href="https://www.raspberrypi.com/news/mechatronics-engineering-as-a-form-of-artistic-expression/">Raspberry Pi</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Revopoint MIRACO 2: 6 Scanning Modes and Standalone Operation</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/c418eeffbfe8</link><guid isPermaLink="false">eink-c418eeffbfe8-en</guid><pubDate>Thu, 19 Nov 2026 13:03:36 GMT</pubDate><category>maker</category><description><![CDATA[Revopoint's MIRACO 2 offers six 3D scanning modes, including blue lasers and infrared, with a touchscreen and battery.]]></description><content:encoded><![CDATA[<p><strong>Revopoint's MIRACO 2 offers six 3D scanning modes, including blue lasers and infrared, with a touchscreen and battery.</strong></p><p>Revopoint has launched a crowdfunding campaign for the <strong>MIRACO 2</strong>, a second-generation all-in-one portable 3D scanner. The device features six scanning modes, combining blue lasers, infrared structured light, and photogrammetric metrology. According to the manufacturer, this versatility allows capturing both fine details and large objects without switching tools.</p><p>The available modes include crossed or parallel blue laser lines for dark or reflective objects, a single-line blue laser mode for hard-to-reach areas, and an infrared structured light scan at 30 frames per second. A hybrid mode combines structured light and VCSEL (vertical-cavity surface-emitting laser) to reach 10 million points per second. Precision varies by mode: up to 0.02 mm in multi-line laser mode, with a volumetric accuracy of 0.02×0.05 mm along the object's length.</p><p>The MIRACO 2 operates standalone with a 10,000mAh battery, but advanced features require a PC connection: color tracking, markerless scanning in laser line mode, smart segmentation, automatic cropping, and GPU-accelerated calculations for up to 120 frames per second. An optional photogrammetric module extends the scan range from 10×10×10 mm to 5×5×5 m. First units are expected in January 2027, with early bird rewards starting at $1,495.
        <p>Source : <a href="https://www.hackster.io/news/revopoint-s-new-miraco-2-3d-scanner-comes-with-six-scanning-modes-standalone-operation-a736af31a2a9">Hackster</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>ZFS in a Homelab: 5 Myths You Should Ignore</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/3f0adc4fcb31</link><guid isPermaLink="false">eink-3f0adc4fcb31-en</guid><pubDate>Thu, 19 Nov 2026 12:30:15 GMT</pubDate><category>maker</category><media:content url="https://static0.howtogeekimages.com/wordpress/wp-content/uploads/wm/2025/11/a-beelink-mini-pc-and-terramaster-nas-together-on-a-shelf.jpg" medium="image"/><description><![CDATA[Using ZFS in a homelab is often discouraged by exaggerated hardware requirements.]]></description><content:encoded><![CDATA[<p><img src="https://static0.howtogeekimages.com/wordpress/wp-content/uploads/wm/2025/11/a-beelink-mini-pc-and-terramaster-nas-together-on-a-shelf.jpg" alt=""></p><p><strong>Using ZFS in a homelab is often discouraged by exaggerated hardware requirements.</strong></p><p>ZFS, a filesystem praised for its data integrity and reliability, is often surrounded by hardware recommendations that discourage its use in homelabs. However, many of these claims are more myth than necessity.</p> <h2>Overstated hardware requirements</h2> <p>One of the most persistent myths is the insistence on using ECC (Error-Correcting Code) memory to run ZFS. <strong>ECC memory detects and corrects bit errors in RAM</strong>, which can improve data reliability, especially on critical servers. However, ZFS is not the only filesystem vulnerable to memory corruption: ext4 or NTFS face similar risks. While the absence of ECC slightly increases the chances of errors, <strong>these risks remain manageable with reliable backups</strong>. An old desktop with standard RAM can therefore run ZFS without issues.</p> <p>Another widespread myth is the <strong>1GB of RAM per 1TB of storage</strong> rule. This recommendation, often linked to ZFS’s deduplication feature, is largely exaggerated. Deduplication, which stores identical data blocks only once, can indeed consume significant memory, but it is not essential for homelabs. Recent versions of OpenZFS (since 2.3) include mechanisms like <strong>Fast Dedup</strong>, which limit the memory footprint of this feature. For a simple home server, 4 to 8GB of RAM is usually sufficient, and lighter setups can work with less.</p> <p>Finally, ZFS’s Adaptive Replacement Cache (ARC), which uses RAM to speed up data access, is sometimes misunderstood. A large ARC does not indicate a lack of memory but rather an optimal use of available resources. The cache can be freed if other applications require it.</p>
        <p>Source : <a href="https://www.howtogeek.com/i-use-zfs-in-my-homelab-and-i-wish-these-5-myths-would-go-away/">How-To Geek</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>SwitchBot Hub 4: Smart Home Hub with Screen and AI Voice Assistant</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/2a471c270a5a</link><guid isPermaLink="false">eink-2a471c270a5a-en</guid><pubDate>Thu, 19 Nov 2026 12:23:15 GMT</pubDate><category>Dashboards</category><description><![CDATA[SwitchBot unveils the Hub 4, a smart home hub featuring a touchscreen, an AI voice assistant, and offline infrared control.]]></description><content:encoded><![CDATA[<p><strong>SwitchBot unveils the Hub 4, a smart home hub featuring a touchscreen, an AI voice assistant, and offline infrared control.</strong></p><p>SwitchBot has launched the Hub 4, a fourth-generation smart home hub designed to centralize control of connected devices. The device features a 2.8-inch color touchscreen, physical rotary controls, and an AI voice assistant called <strong>Hello Kata</strong>.</p><p>The Hub 4 includes an <strong>infrared blaster</strong> capable of operating over 101,000 compatible appliances (air conditioners, TVs, projectors) offline once devices are configured. It supports <strong>Matter 1.3</strong>, enabling integration with third-party platforms like Apple Home, while managing up to 30 sub-devices — including three built-in sensors (motion, temperature, humidity). The touchscreen interface uses expressive <strong>animated eyes</strong> for a more engaging experience, and a rotary knob simplifies navigation. Four configurable shortcuts allow quick access to custom <strong>scenes</strong>.</p><p>The <strong>Hello Kata</strong> voice assistant, powered by OpenAI’s <strong>GPT-6</strong> model, responds to natural language commands but requires an internet connection. The Hub 4 also supports streaming devices from <strong>Android TV, Apple TV, and Fire TV</strong>.</p><p>Available for pre-order on SwitchBot’s website at $119.98 (or $95.99 with the code <strong>BCFMX20</strong>), the first units are scheduled for early December 2026 delivery.
        <p>Source : <a href="https://www.hackster.io/news/switchbot-unveils-the-hub-4-a-matter-bridge-and-smart-home-dashboard-with-emotive-user-interface-41e02534113e">Hackster</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Apple HomeView: A 2026 Anti-Repair Display</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/70b1881d86e3</link><guid isPermaLink="false">eink-70b1881d86e3-en</guid><pubDate>Thu, 19 Nov 2026 12:00:06 GMT</pubDate><category>maker</category><media:content url="https://hackaday.com/wp-content/uploads/2026/11/apple_homeview_inside_ifixit_youtube.jpg" medium="image"/><description><![CDATA[Apple’s HomeView teardown exposes design choices that severely hinder repairs, despite the company’s repairability promises.]]></description><content:encoded><![CDATA[<p><img src="https://hackaday.com/wp-content/uploads/2026/11/apple_homeview_inside_ifixit_youtube.jpg" alt=""></p><p><strong>Apple’s HomeView teardown exposes design choices that severely hinder repairs, despite the company’s repairability promises.</strong></p><p>The <strong>iFixit</strong> teardown of Apple’s new <strong>HomeView</strong> display reveals major repair hurdles despite the company’s repairability claims. The device uses ultra-strong adhesive to secure the screen and speakers, making removal difficult. The mainboard’s rear metal can is glued in place with thermal compound, and the speaker connectors on the board crumble upon first attempt to unplug them, even after cleaning with isopropyl alcohol (IPA).</p><p>Disassembling the audio modules proves particularly challenging: one speaker is nearly impossible to reattach after separation, while the subwoofer in the base version is unusable without harsh solvents or invasive tools. The wall-mounted version, though simpler, still suffers from limited screw access and pervasive adhesive.</p><p>These design choices, described by iFixit as <strong>“actively resisting repairs”</strong>, suggest a deliberate effort to make the device hard to fix. Additionally, the HomeView’s absence from the EU market may stem from Siri’s privacy concerns, but could also conflict with right-to-repair regulations.</p>
        <p>Source : <a href="https://hackaday.com/2026/11/19/how-apples-homeview-actively-resists-repairs/">Hackaday</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>RetroLine: A Bluetooth Adapter to Link Old Landline Phones to Smartphones</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/d1e5bbce77e0</link><guid isPermaLink="false">eink-d1e5bbce77e0-en</guid><pubDate>Thu, 19 Nov 2026 11:34:03 GMT</pubDate><category>maker</category><description><![CDATA[An Australian device turns analog landline phones into Bluetooth peripherals for smartphones.]]></description><content:encoded><![CDATA[<p><strong>An Australian device turns analog landline phones into Bluetooth peripherals for smartphones.</strong></p><p><strong>Indigo Telephony</strong>, an Australian startup, has unveiled the <strong>RetroLine</strong>, a Bluetooth adapter that connects old analog landline phones to smartphones. The microcontroller-based device plugs between the handset and the existing phone line, acting as a miniature private branch exchange (PBX). It provides a dial tone when the handset is lifted and converts DTMF (touch-tone) or pulse dialing into numbers dialed from the paired smartphone.</p><p>The RetroLine enables calls to be made and received while offering advanced features like call recording to a microSD card, customizable hold music, and a web dashboard to manage ringtones, greetings, and call automations. Analog audio from the connected phone is digitized and transmitted over high-definition Bluetooth, with manual volume controls for the microphone, earpiece, and hold music.</p><p>The device can connect to the local Wi-Fi network to access online voice services or voice assistants, or operate independently by hosting its own hotspot. Future updates include VoIP support and a "walkie-talkie" mode for direct unit-to-unit communication. Priced at <strong>AU$160 (around $107)</strong>, the RetroLine is currently funding on Kickstarter, with hardware expected to ship in <strong>April 2027</strong>. As with all crowdfunding campaigns, fulfillment is not guaranteed, especially since this is the company’s first product.
        <p>Source : <a href="https://www.hackster.io/news/indigo-telephony-s-retroline-brings-your-old-analog-landline-back-to-life-via-your-cellphone-4df9eb7d4a87">Hackster</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>An Open-Source E-Ink Wi-Fi Sign to Share Passwords Without the Hassle</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/4dfc7602755c</link><guid isPermaLink="false">eink-4dfc7602755c-en</guid><pubDate>Thu, 19 Nov 2026 11:00:15 GMT</pubDate><category>displays</category><media:content url="https://static0.howtogeekimages.com/wordpress/wp-content/uploads/wm/2026/11/wifi-sign-thanksgiving-front.jpg" medium="image"/><description><![CDATA[A DIY project using a Raspberry Pi and an e-ink display to show a Wi-Fi QR code, sparing guests from typing a complex password.]]></description><content:encoded><![CDATA[<p><img src="https://static0.howtogeekimages.com/wordpress/wp-content/uploads/wm/2026/11/wifi-sign-thanksgiving-front.jpg" alt=""></p><p><strong>A DIY project using a Raspberry Pi and an e-ink display to show a Wi-Fi QR code, sparing guests from typing a complex password.</strong></p><p>To avoid repeating a long, complex Wi-Fi password to every visitor, a maker built an electronic sign using an <strong>e-ink display</strong> powered by a Raspberry Pi. The device shows a QR code containing the guest network credentials, allowing visitors to connect by scanning it—no need to memorize or type a security key.</p> <p>The e-ink screen, chosen for its low power consumption and static display without power, is paired with a <strong>Raspberry Pi</strong> and software generated by the AI <strong>Claude</strong> in under 30 minutes. The project relies on a QR code encoding Wi-Fi details in the format WIFI:T:WPA;S:mynetwork;P:mypass;;, where S is the network name (SSID) and P the password. The code is customizable and can include seasonal messages or varied graphics.</p> <p>The maker used a Waveshare 7.3-inch color e-ink display, now replaced by a Spectra 6 model, but notes that monochrome alternatives (like a 2.13-inch screen under $30) work for this purpose. The project is open source and available on GitHub, with options to hide or display the password as needed. Automatically rotating Wi-Fi credentials every 24 hours is also possible for added security.
        <p>Source : <a href="https://www.howtogeek.com/i-built-an-open-source-e-ink-wi-fi-sign-so-i-could-finally-give-my-guest-network-a-real-password/">How-To Geek</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Open-Source I2C TPM 2.0 Security Module for Raspberry Pi Launches Crowdfunding</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/3d00fbcf6168</link><guid isPermaLink="false">eink-3d00fbcf6168-en</guid><pubDate>Thu, 19 Nov 2026 09:43:17 GMT</pubDate><category>maker</category><media:content url="https://www.cnx-software.com/wp-content/uploads/2026/11/ANAVA-I2C-TPM-2.0-security-module-720x374.jpg" medium="image"/><description><![CDATA[An open-source I2C TPM 2.0 security module for Raspberry Pi, based on an Infineon chip and TCG-certified, is now crowdfunding.]]></description><content:encoded><![CDATA[<p><img src="https://www.cnx-software.com/wp-content/uploads/2026/11/ANAVA-I2C-TPM-2.0-security-module-720x374.jpg" alt=""></p><p><strong>An open-source I2C TPM 2.0 security module for Raspberry Pi, based on an Infineon chip and TCG-certified, is now crowdfunding.</strong></p><p>An open-source I2C TPM 2.0 security module, the <strong>ANAVI I2C TPM 2.0</strong>, is designed for Raspberry Pi single-board computers via an I2C interface. It uses the <strong>Infineon Optiga SLB 9673</strong> chip, a TPM 2.0 module compliant with Trusted Computing Group (TCG) specifications. This component enhances security for Raspberry Pi projects by providing encryption, secure key storage, and hardware-based random number generation (TRNG), meeting NIST SP 800-90A/B standards.</p><p>The module comes in two mechanical configurations: a horizontal version and a vertical version, both featuring a 5-pin (2.54 mm pitch) female header for I2C, reset, 3.3V, and GND connections. Hardware design files, created with KiCad, are released under the Creative Commons Attribution-ShareAlike 4.0 license, allowing modifications and commercial use. The project is OSHWA-certified (BG000154).</p><p>Software integration is streamlined with a device tree overlay (tpm-slb9673.dtbo) included in the Raspberry Pi Linux kernel, ensuring compatibility with Raspberry Pi OS and other compatible Linux distributions. A step-by-step tutorial demonstrates how to set up an encrypted partition using the module.</p><p>Shipments are scheduled to begin on March 18, 2027. The module is available on Crowd Supply for $20 (horizontal) or $21 (vertical), with shipping costs varying by destination.
        <p>Source : <a href="https://www.cnx-software.com/2026/11/19/raspberry-pi-sbcs-get-an-open-source-hardware-i2c-tpm-2-0-security-module/">CNX Software</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Mesh2Cad Converts 3D Mesh to CAD File</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/3be796f275d0</link><guid isPermaLink="false">eink-3be796f275d0-en</guid><pubDate>Thu, 19 Nov 2026 09:00:52 GMT</pubDate><category>maker</category><media:content url="https://hackaday.com/wp-content/uploads/2026/11/mesh2cad_before_after.png" medium="image"/><description><![CDATA[A Python tool automates STL-to-CAD conversion for 3D printing or machining workflows.]]></description><content:encoded><![CDATA[<p><img src="https://hackaday.com/wp-content/uploads/2026/11/mesh2cad_before_after.png" alt=""></p><p><strong>A Python tool automates STL-to-CAD conversion for 3D printing or machining workflows.</strong></p><p>STL files are convenient for 3D printing but often fall short when precise modifications are needed in a CAD environment. <strong>Mesh2Cad</strong>, created by <strong>Emir Mrkaljevic</strong>, bridges this gap by automatically converting 3D meshes into usable CAD models. Unlike other tools, it doesn’t lock users into a single format: the software generates scripts for <strong>FreeCAD</strong>, <strong>CadQuery</strong>, or <strong>OpenSCAD</strong>, as well as STEP files, depending on the workflow.</p> <p>Installation is straightforward via pip install mesh2cad, but the code is also available on <strong>GitHub</strong> for manual review. The scripts include built-in validation: the conversion only proceeds if the generated CAD model matches the original mesh within a set tolerance. This reduces errors compared to manual reconstruction, especially for simple geometric shapes like boxes, brackets, or mechanical parts with holes and chamfers.</p> <p>That said, <strong>Mesh2Cad</strong> isn’t designed for organic shapes or complex curves common in realistic 3D scans. For those cases, manual editing remains necessary. The tool targets users who need to tweak STL files but lack the time or tools to rebuild them from scratch in a CAD program.</p>
        <p>Source : <a href="https://hackaday.com/2026/11/19/mesh2cad-automates-re-modelling/">Hackaday</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
<item><title>Dreamcart: Playing N64 Cartridges on a Dreamcast</title><link>https://nbbou81000.github.io/trmnl-leaderboard/?lang=en#eink/b0ce4b53a6ee</link><guid isPermaLink="false">eink-b0ce4b53a6ee-en</guid><pubDate>Thu, 19 Nov 2026 08:29:34 GMT</pubDate><category>maker</category><media:content url="https://korben.info/dreamcast-cartouches-n64/dreamcast-cartouches-n64-1.jpg" medium="image"/><description><![CDATA[A hacker enables Nintendo 64 cartridges to run on a Dreamcast using a custom reader and a dynamic recompiler.]]></description><content:encoded><![CDATA[<p><img src="https://korben.info/dreamcast-cartouches-n64/dreamcast-cartouches-n64-1.jpg" alt=""></p><p><strong>A hacker enables Nintendo 64 cartridges to run on a Dreamcast using a custom reader and a dynamic recompiler.</strong></p><p>The <strong>Dreamcart</strong> project, created by hacker <strong>Throaty Mumbo</strong>, allows inserting a Nintendo 64 cartridge into a Dreamcast’s controller port using a custom reader. This reader leverages the Maple bus, already used for Dreamcast peripherals, and relies on an <strong>RP2350</strong> chip. A modified version also supports Super Nintendo and Mega Drive cartridges.</p><p>To overcome the Dreamcast’s limited processing power, Dreamcart employs a <strong>dynamic recompiler</strong>, inspired by the <strong>N64Recomp</strong> project. This software translates N64 game code into instructions the Dreamcast can execute, avoiding the sluggishness of traditional emulation. Pre-compiled games can be stored on an SD card for faster startup.</p><p>Tested titles that run well include <strong>Super Mario 64</strong>, <strong>Mario Kart 64</strong>, <strong>Kirby 64</strong>, <strong>Rayman 2</strong>, <strong>Tony Hawk’s Pro Skater 2</strong>, <strong>Cruis’n USA</strong>, and <strong>Elmo’s Number Journey</strong>. <strong>Pokémon Stadium 2</strong> runs with slowdowns, while <strong>Donkey Kong 64</strong> fails to start. For Super Nintendo and Mega Drive cartridges, modified versions of the <strong>Super Famicast</strong> and <strong>Genesis Plus DC</strong> emulators handle compatibility.</p><p>The project remains a prototype requiring technical steps: burning <strong>dcload-ip</strong> onto a disc, sending the program from a PC via <strong>dc-tool-ip</strong> (using the Dreamcast’s Ethernet adapter), assembling the reader from available schematics, and flashing the firmware with a debug probe. The creator advises testing with a low-value game first and saving progress, though save files are tied to the physical cartridges.
        <p>Source : <a href="https://korben.info/dreamcast-cartouches-n64.html?utm_source=rss&amp;utm_medium=feed&amp;utm_campaign=flux-principal">Korben</a></p>
        <p><em>Article written by an AI (Mistral) from the cited source: check important details at the source.</em></p>]]></content:encoded></item>
</channel></rss>
