news

Qwen 3.8 Launches, OpenAI Cuts Context, Minecraft Uses SDL3

19 Jul 2026
Qwen 3.8
Qwen 3.8

Qwen 3.8 is launching soon as an open-weight AI model boasting 2.4T parameters. The continuously evolving system is considered one of the most powerful available today, ranking second only to Fable 5 among leading frontier models.

Comments
Show HN: I replaced a $120k bowling center system with $1,600 in ESP32s

A developer successfully replaced a $120,000 bowling center control system with a custom $1,600 solution built using ESP32 microcontrollers. This project sparked widespread discussion about the broader potential of retrofitting legacy industrial and automotive equipment with modern, low-cost embedded technologies. Community members highlighted both the technical viability and the regulatory challenges of creating niche aftermarket solutions for outdated machinery.

Comments
Claude Code uses Bun written in Rust now

Bun creator Jarred Sumner announced that Claude Code version 2.1.181 and later now use the runtime's rewritten Rust port. This update makes the AI coding tool 10% faster on startup.

Comments
What I learned selling 2,500 MIDI recorders: Hardware is not so hard

Chip Weinberger reflects on his experience designing and selling 2,500 Jamcorder MIDI recorders to share key lessons for aspiring product creators. He debunks the common tech industry myth that building physical products is overwhelmingly difficult, proving that hardware development is actually quite manageable.

Comments
Minecraft: Java Edition now uses SDL3
Minecraft: Java Edition now uses SDL3

Minecraft Java Edition now utilizes SDL3 in the newly released 26.3 Snapshot 4 update. This backend transition modernizes the game's underlying multimedia and hardware abstraction layer for improved cross-platform support.

Comments
Transcribe.cpp

The Transcribe.cpp project is a new initiative designed to simplify audio transcription while providing an interactive sandbox for tech enthusiasts. By allowing users to play around with the technology easily, it aims to make complex code integrations both straightforward and fun.

Comments
NYC may require landlords and realtors to disclose the use of AI in listings
NYC may require landlords and realtors to disclose the use of AI in listings

New York City Mayor Mamdani has announced proposed regulations requiring landlords and real estate agents to disclose the use of artificial intelligence in property listings. This initiative is designed to increase transparency and protect tenants navigating the city's housing market.

Comments
I joined the IndieWeb, here's what I learned
I joined the IndieWeb, here's what I learned

Prompted by curiosity after noticing the IndieWeb tag across various blogs and Mastodon posts, the author explores the decentralized web movement. By joining the community, the writer shares their firsthand experience and key takeaways about taking ownership of personal digital content.

Comments
Bananas sprout in Rayleigh Garden UK after 15 years
Bananas sprout in Rayleigh Garden UK after 15 years

After fifteen years of effort, a couple in Rayleigh, Essex, has successfully grown bananas in their UK garden. Emma and Steve Stav achieved this rare feat by cultivating an impressive collection of 200 banana plants on their property.

Comments
Better and Cheaper Than IPTV
Better and Cheaper Than IPTV

A new tool named Castor offers a cheaper alternative to traditional IPTV by allowing users to cast web-based videos directly to their televisions. By simply pointing the software at any web page, it automatically extracts, transcodes, and streams the video in real time while burning in subtitles.

Comments
Mathematicians still don't know the fastest way to multiply numbers
Mathematicians still don't know the fastest way to multiply numbers

A 23-year-old student has overturned an ancient mathematical conjecture regarding the efficiency of number multiplication. Despite this recent breakthrough, mathematicians still do not know the absolute fastest way to multiply numbers.

Comments
OpenAI reduces Codex Model Context Size from 372k to 272k
OpenAI reduces Codex Model Context Size from 372k to 272k

OpenAI has reduced the context window size of its Codex model from 372k to 272k tokens. The company is backporting this change to the stable 0.144 release alongside refreshed GPT-5.6 model instructions and metadata. These updates ensure feature consistency between stable clients and the latest public main branch.

Comments
Blender 5.2 LTS
Blender 5.2 LTS

Blender has released version 5.2, marking the first long-term support update in the 5.x series. This update guarantees ongoing maintenance and stability for users until July 2028.

Comments
C64 Basic Dungeon Crawler: Goblin Attack (C64 Basic Part 8)
C64 Basic Dungeon Crawler: Goblin Attack (C64 Basic Part 8)

The eighth installment of a Commodore 64 BASIC programming series details the development of a new dungeon crawler called Goblin Attack. The game features redefinable keys, chasing enemies, and a d20 combat system. The tutorial also covers the technical challenge of fixing a critical memory bug that caused the game to silently overwrite its own code.

Comments
Classic Amiga titles, free to download

A comprehensive archive of classic Amiga public domain software, featuring historic collections from 17-Bit and Fred Fish, is now available for free download. Beyond offering retro computing nostalgia, the extensive library provides valuable historical insights into early software development and the foundational evolution of standardized data formats.

Comments
Cagire: Live Coding in Forth
Cagire: Live Coding in Forth

Cagire is a terminal-based step sequencer that enables live music production and coding using the Forth programming language. By assigning a Forth script to each step, the single binary integrates synthesis, sampling, and support for MIDI and OSC to provide a unique command-line environment for musicians and developers.

Comments
HMD Touch 4G
HMD Touch 4G

HMD has unveiled the Touch 4G, a compact hybrid phone featuring a touchscreen and up to 30 hours of battery life. The device bridges basic and smart phone capabilities with WiFi, Cloud Phone Service, and an Express Chat App for text, audio, and video calls.

Comments
I burned all my tokens researching how to save tokens
I burned all my tokens researching how to save tokens

A developer maxed out their Claude subscription limit in thirty minutes while researching methods to reduce AI token consumption. Through this process, they built a custom deep research pipeline that uses shared memory and assigns distinct roles across three existing AI subscriptions. The author outlines how readers can replicate this cost-effective system using services they already pay for.

Comments
AI Mania Is Eviscerating Global Decision-Making

Reader discussions surrounding the hype of artificial intelligence reveal a stark divide in user experiences, with some praising large language models as autonomous coding superpowers while others dismiss them as unreliable tools requiring constant micro-management. This polarization has sparked intense debate over whether the technology's effectiveness is merely a reflection of a user's underlying technical competence, highlighting a growing disconnect between the promised technological singularity and the practical realities of AI integration.

Comments
HomeLab #1: MikroTik as a Home Router
HomeLab #1: MikroTik as a Home Router

The latest HomeLab series details the experience of configuring a MikroTik device to serve as a primary home router. Although the author cautions readers about the steep learning curve and warns to proceed at their own risk, the guide highlights that the advanced networking features make the complex setup highly rewarding.

Comments
AI advice made people less accurate but more confident – sudy
AI advice made people less accurate but more confident – sudy

A recent study reveals that relying on AI advice makes individuals less accurate but significantly more confident in their responses. When provided with AI guidance, people's willingness to admit uncertainty plummeted from 44% to just 3%. As a result, their overall task accuracy dropped sharply from 27% to 9% despite their heightened confidence.

Comments
Dupes took over the world
Dupes took over the world

The Verge's Mia Sato explores the massive cultural and economic impact of product dupes, which have become a dominant force in modern consumer markets. The article breaks down why shoppers are increasingly flocking to affordable alternatives and examines the highly profitable business models driving this global trend.

Comments
The death and rebirth of my home server

Tech blogger Steen shares a personal account of rebuilding their home server following a complete hardware failure. The article outlines the technical challenges of the data migration process and the final configuration of the newly resurrected self-hosted setup.

Comments
Moonshot AI suspends new subscriptions due to Kimi K3 demand
Moonshot AI suspends new subscriptions due to Kimi K3 demand

Moonshot AI has temporarily suspended new subscriptions for its Kimi K3 model. An unexpected surge in user demand over the past 48 hours has pushed the company's GPU infrastructure to its maximum capacity. The pause is intended to ensure a stable experience for existing subscribers while the company manages the overwhelming traffic.

Comments
UnifiedIR for Julia
UnifiedIR for Julia

The official GitHub repository for the Julia programming language is highlighting new development efforts focused on implementing a UnifiedIR. Developers are encouraged to create an account and contribute to the ongoing compiler infrastructure improvements. This initiative represents a significant step in the language's continuous development and optimization.

Comments
A Visual Catalog of Retro Macintosh Software

A newly released visual catalog showcases an extensive collection of vintage Macintosh software and classic applications. This online archive allows users to explore the early graphical interfaces of Apple computers, serving as both a nostalgic trip for longtime users and a historical resource for tech enthusiasts.

Comments
A new Intel Itanium (IA-64) emulator that boots Windows

A newly developed emulator can now boot Windows on the discontinued Intel Itanium architecture, albeit at speeds comparable to a 1990s 486 processor on modern hardware. While this milestone is a notable technical achievement, community discussions highlight that Itanium's VLIW design fundamentally failed for general-purpose computing. Specifically, the architecture's reliance on predictable memory access and complex compiler optimizations could not efficiently handle the unpredictable multitasking and virtualization demands of modern workloads.

Comments
Scrying the AMD GFX1250 LLVM Tea Leaves
Scrying the AMD GFX1250 LLVM Tea Leaves

AMD is preparing to unveil its new MI400 series Datacenter Accelerators at the upcoming Advancing AI event. In anticipation of the official reveal, developers are analyzing recent LLVM code commits to uncover technical details about the underlying GFX1250 architecture.

Comments
The Last MPEG-4 Visual Patent Has Expired
The Last MPEG-4 Visual Patent Has Expired

The final patent covering the MPEG-4 Part 2 visual standard has officially expired, freeing the widely used video format from all remaining licensing restrictions. This significant milestone allows developers and companies to implement the technology without worrying about patent royalties.

Comments
Making Software: How to make a font

A recent article explores the technical mechanics of font encoding, rendering, and typography fundamentals such as kerning and ligatures rather than providing a practical design tutorial. The explainer was praised by readers for its clear breakdown of complex typographic concepts and the technical distinctions between different font formats. It also sparked community discussions regarding custom font creation tools and the potential computational performance impacts of highly complex character designs.

Comments