Published on August 02, 2026 at 18:04 CEST (UTC+2)
Twenty Years of RISC OS Open (82 points by AlexeyBrin)
RISC OS Open marks its twentieth anniversary with a retrospective of how the organization was founded in 2006 to open-source the proprietary RISC OS operating system. The article traces the project through yearly milestones, including the first public source releases, community-contributed builds, and nightly autobuilds. It credits the community of developers, testers, and contributors for the platform's improved health over the past two decades.
F*: A general-purpose proof-oriented programming language (61 points by ducktective)
F* is a general-purpose, proof-oriented programming language that combines dependent types with SMT solving and tactic-based interactive theorem proving. It supports both purely functional and effectful programming, and programs can be extracted to OCaml, F#, C, Wasm, or assembly. The language is open source, actively developed by Microsoft Research, Inria, and the community, and includes tutorials and online materials for learning.
Meshdiff – visually compare two STL versions in the browser, client-side (114 points by projscope)
Meshdiff is a browser-based tool for visually comparing 3D model versions in formats like STL, 3MF, and OBJ. It runs entirely client-side, requiring JavaScript, and is designed to help users diff 3D files without uploading them to a server. This makes it useful for version control and collaboration in 3D printing and CAD workflows.
Show HN: Bor – Open-source policy management for Linux desktops (121 points by eniac111)
Bor is an open-source policy management tool for Linux desktops, and the v0.8.0 release adds support for Thunderbird, Microsoft Edge for Business, and Firewalld zone policies. The release also includes a full web UI overhaul, finer-grained role-based access control, and a security hardening pass. The agent writes expected managed-policy files for each application and includes a tamper watcher that detects and reverts external changes.
Folding Paper Globes (60 points by dango2506)
Folding Paper Globes is a website offering free printable paper globe templates. It includes globes of solar system bodies such as the Sun, Moon, Mars, and Jupiter, as well as various Earth visualizations like land temperature, ocean life, and population density. The collection also features data globes showing earthquakes, volcanoes, tectonic plates, and meteorite landings for educational purposes.
Fasttracker II clone in C using SDL 2 (38 points by andsoitis)
Fasttracker II clone is a portable reimplementation of the classic Fasttracker II music tracker, written in C using SDL 2. The author provides downloads for Windows, macOS, and Linux, along with source code and compilation instructions. The page also includes platform-specific notes about keyboard shortcuts, multi-monitor setups, and macOS security settings.
Artificial Intelligence: Ars Notoria and the Promise of Instant Knowledge (82 points by jruohonen)
Ars Notoria and the Promise of Instant Knowledge examines a medieval magical manuscript that claimed to rapidly transmit advanced knowledge through diagrams, incantations, and rituals. The essay notes that while church authorities condemned the work, many manuscripts survived, showing its enduring appeal. It draws a historical parallel to modern desires for instant learning and, as the title suggests, artificial intelligence.
A Rant About “Technology” (2005) (74 points by jamesgill)
Ursula K. Le Guin's rant about "technology" pushes back against the idea that her science fiction avoids technology because it is not "hard" SF. She argues that all genuine science fiction includes technological content, including technologies of mind, society, and culture. Le Guin criticizes the narrow equation of technology with machinery and engineering, emphasizing broader definitions of how human worlds are built.
Show HN: Fuse – statically typed functional programming language (53 points by the_unproven)
Fuse is a statically typed, purely functional programming language with higher-kinded types and ad-hoc polymorphism. It compiles through the GRIN whole-program optimizer to LLVM-generated native code. The language features System F–based type checking, bidirectional type inference, pattern matching, and a syntax inspired by Rust, Python, Scala, and Haskell.
Go 1.27 Interactive Tour (297 points by Hixon10)
Go 1.27 Interactive Tour is a hands-on, runnable walkthrough of the upcoming Go 1.27 release. It was created by VictoriaMetrics, continuing a series originally started by Anton Zhiyanov for Go 1.22 through 1.26. The tour links to official release notes, proposals, commits, and authors, making it a practical complement to the more formal documentation.
The "instant knowledge" promise is a recurring cultural pattern that AI now embodies.
The Ars Notoria essay and Le Guin's rant both critique the seduction of effortless access to knowledge or technology. Modern LLMs and AI assistants frequently make the same promise: instant answers, rapid mastery, and minimal effort.
Why it matters: AI/ML systems that optimize for fast, confident outputs can reinforce shallow understanding and over-trust in generated answers, just as the Ars notoria offered apparent mastery without genuine study.
Implications: Developers should build AI tools that surface uncertainty, support verification, and encourage deeper engagement with source material rather than merely delivering convincing text.
Proof-oriented programming and formal verification are becoming more relevant to AI/ML reliability.
F shows how dependent types, SMT solving, and tactic-based proving can be used to build verified software. As AI-generated code and autonomous systems grow, the need for mathematically grounded guarantees becomes more pressing.
Why it matters: Machine learning outputs are probabilistic, but downstream systems often require correctness and safety guarantees. Proof-oriented languages can verify properties of code paths, data transformations, and decision logic.
Implications:* AI/ML engineering should explore integrating formal verification tools into model-serving pipelines, especially for safety-critical applications where errors are costly.
Client-side and browser-based tools are democratizing access to specialized technical workflows.
Meshdiff performs complex 3D geometry diffing entirely in the browser, and the Go 1.27 interactive tour brings runnable examples to users without local setup. This mirrors broader trends in AI/ML: WebAssembly, on-device inference, and interactive notebooks lower the barrier to experimentation.
Why it matters: Privacy-preserving AI can run locally in browsers or edge devices, while interactive tours and demos make machine learning concepts more approachable.
Implications: Teams should invest in client-side tooling, interactive documentation, and in-browser model demos to improve adoption, transparency, and user trust.
Open-source longevity and community governance remain critical infrastructure for AI/ML.
RISC OS Open's twenty-year journey illustrates how sustained community effort keeps projects alive. Bor similarly represents a community-driven governance layer for desktop policy. AI/ML itself depends on open-source frameworks, datasets, and toolchains that require ongoing maintenance.
Why it matters: Without sustainable funding, contributor pipelines, and governance, critical AI infrastructure can become unmaintained or concentrated in a few corporations.
Implications: Organizations should support maintainers, establish clear governance models, and treat open-source AI dependencies as core infrastructure deserving investment and resilience planning.
Functional and statically typed languages offer advantages for complex AI/ML systems.
Fuse and F emphasize purity, strong typing, higher-kinded abstractions, and formal reasoning. These features help manage complexity in data pipelines, model definitions, and code generation.
Why it matters: AI/ML codebases often involve intricate transformations and difficult-to-trace side effects. Purely functional design can make behavior more predictable and testable.
Implications:* AI/ML developers should consider functional programming concepts and typed DSLs for critical components such as data preprocessing, feature engineering, and model orchestration.
Interactive, example-driven learning is becoming the standard for programming and AI/ML education.
The Go 1.27 tour continues a successful pattern of hands-on walkthroughs with runnable examples. Interactive notebooks and browser-based labs have already transformed machine learning education.
Why it matters: As AI/ML frameworks evolve rapidly, static documentation is insufficient. Users need to experiment directly with code and see behavior in real time.
Implications: Tool builders and educators should prioritize interactive tutorials, live examples, and low-friction sandboxes when introducing new models, APIs, or techniques.
Policy management is expanding to cover software endpoints, including AI assistants and browser-based agents.
Bor's addition of policy types for Thunderbird, Edge, and Firewalld reflects a broader need to manage application behavior across Linux desktops. As AI-powered assistants and agents become embedded in browsers and operating systems, similar policy controls will be necessary to govern permissions, data access, and automation.
Why it matters: Enterprises will want to restrict what AI tools can read, write, or execute on managed devices. Without policy management, AI agents could become security and compliance risks.
Implications: Open-source policy systems like Bor could evolve to include AI-specific rule sets, enabling organizations to enforce safe and accountable use of local AI deployments.
Analysis generated by deepseek-reasoner