Radiant Computer

Original Article ↗ Hacker News Discussion ↗

Accessibility & UI Concerns

  • Multiple commenters worry a “clean-slate” OS will neglect accessibility, especially for blind users and screen readers.
  • Some argue accessibility can be layered on later if the GUI exposes proper metadata and state; others counter that good screen readers need deep integration and stable element identity, which is hard to retrofit.
  • There’s also a broader notion of “accessibility” as approachability for non-experts, which the project claims to care about.
  • Visual design of the site (low-contrast grey-on-grey) is criticized as unfriendly and “loading-screen-like,” reinforcing concerns.

AI‑Native OS: Interest vs Distrust

  • The “AI-native” positioning is polarizing. Some are intrigued by an OS designed from the ground up with local models and structured app metadata, seeing it as a fix for today’s bolted‑on AI.
  • Others view AI as inherently untrustworthy or at odds with the project’s stated values of human-centric, simple, tractable systems.
  • Concerns include ethical training data, hardware practicality (FPGA vs GPU for local LLMs), and fear that AI integration undermines the anti‑surveillance, anti‑social‑media ethos.

Clean‑Slate Stack: Hardware, OS, Language

  • Ambition: custom RISC‑V hardware, exokernel‑like/single‑address‑space OS, new systems language (Radiance/Rʹ), tight HW–SW co‑design, no default browser, emphasis on local-first, capability security, REPL‑centric, easily scriptable GUI.
  • Some praise the willingness to rethink everything, comparing it to BeOS, Smalltalk, Rebol-style live systems, or past semantic OS ideas.
  • Others question inventing a new language instead of using Rust/Zig, and criticize the “first principles” rhetoric as vague manifesto rather than grounded design.

Substance, Maturity, and Risk of Vaporware

  • By the project’s own status/log, work is at the compiler/bootstrap stage (Rʹ parser/compiler); no UI, device, or booting OS yet.
  • Many see the site as concept art: sweeping goals (OS, language, hardware, AI) but few concrete artifacts, making people doubt feasibility without large resources.
  • A minority view that as acceptable for an early exploratory experiment, valuing the “north star” vision even if it never ships.

Site Aesthetics and Communication

  • Artwork and copy are widely perceived as AI-generated or “LLM‑ish”: grandiose, repetitive, buzzword-heavy, and more polished than the underlying tech.
  • Some find the retro‑futurist aesthetic inspiring; others say it makes the project feel like vapor, a TV show promotion, or “art school project.”

Compatibility, Networking, and Social Aspects

  • Several note that without browsers or support for mainstream apps (e.g., Discord), a new system is hard to adopt in practice, regardless of technical merit.
  • The “no social networking” principle is divisive: some love escaping social media; others argue social networks are the most valuable part of modern computing.