First screen, captured 9 days ago. Open the video.

Evergreen: A Provably Safe Systems Language

What HN thinks

A provably safe language with not-yet-provable claims, and a forty-comment detour into whether :: is ugly.

Docs and claims need real work
Many found the AI-written README and tutorial shallow or confusing, and said bold claims like "safer than Rust" and "more powerful than C" need benchmarks, comparisons and evidence.
How is this different from existing tools?
Commenters compared it to SPARK, Rust with Verus or Kani, refinement or Liquid types, and asked why a new language beats bolting proofs onto Rust.
Interesting memory model, unclear soundness
The author's "planes" approach to tracking borrows without lifetimes drew skeptical questions about use-after-free, allocations and threads, and some interest as a less restrictive alternative to Rust.

Consensus: Most agree the concept is interesting but under-documented and over-claimed, and that the project needs a spec and a clear comparison with SPARK, Rust and Verus.

Recommended in the comments

Lean (tool): Suggested for building an AI-assisted language on a formally proven calculus, or for writing proofs directly.
Verus (tool): Cited as a way to get proofs on top of standard Rust and its ecosystem.
Kani (tool): Mentioned as an existing option for proving properties of Rust code.
SPARK (tool): Noted as an established language already filling this niche.
Google Technical Writing Courses ↗ (other): Recommended as a style guide for steering LLM-written docs.
Diátaxis ↗ (other): Suggested so the project adds a proper language reference alongside the tutorial.

Summarised by Claude from the comment thread 8 days ago.

Points: 96
Comments: 85
Best rank: #4
Posted: 2026-11-04 by jordan_bonecut
Category: Programming

programming-languages formal-verification systems-programming evergreen contracts webassembly rust

All captures:

Full page

Captured 9 days ago. Click for the whole screenshot.

Looks similar

Screenshot of johnnysswlab.com: Horrible Code, Clean Performance

Horrible Code, Clean Performance (johnnysswlab.com)

112 points by signa11 3 years ago | 89 comments

Screenshot of stork.ai: Apple capped engineers at a $200 daily AI budget

Apple capped engineers at a $200 daily AI budget (stork.ai)

60 points by GloVin 4 days ago | 76 comments

A Stork AI daily newsletter issue led by a report that Apple cut some engineers' daily AI spending cap to $200, plus other AI news items.

apple ai-costs llm-inference tokenization

Screenshot of lograker.app: Adding RSA Support to swift-nio-ssh

Adding RSA Support to swift-nio-ssh (lograker.app)

6 points by moonlighter 1 day ago | 2 comments

Blog post from LogRaker describing how the author forked Apple's swift-nio-ssh to add RSA key support (rsa-sha2-512) for older EC2 instances, and the gotchas involved.

swift-nio-ssh swift ssh rsa

Screenshot of qsoe.net: A QNX-inspired operating system with selectable kernels

A QNX-inspired operating system with selectable kernels (qsoe.net)

34 points by edvinbesic 4 months ago | 4 comments

Screenshot of unfix.dev: Show HN: Green tests are not proof – verifier for C/C++ and firmware fixes

Show HN: Green tests are not proof – verifier for C/C++ and firmware fixes (unfix.dev)

10 points by u-pump 8 days ago | 0 comments

Product page for Unfix, a deterministic verifier that checks whether AI- or human-written C/C++ and firmware bug fixes truly fix bugs by reproducing, applying, and reverting them.

c-cpp firmware ai-coding-agents fix-verification