Zig by Example

Original Article ↗ Hacker News Discussion ↗

Zig programmers are reacting skeptically to the new “Zig by Example” repository, noting that it targets an outdated compiler version, appears AI-generated, and offers little beyond the official docs and more mature learning resources like Ziglings and other community guides. The exchange broadens into whether Zig is worth investing in compared to Rust or C, given its unstable pre‑1.0 status, niche adoption, and lack of edition-style versioning. Some see Zig’s simplicity, toolchain, and C interop as compelling for systems work, while others doubt its long‑term prospects and criticize both its safety model and parts of its community stance on AI.

State of “Zig by Example”

  • Targets Zig 0.14; commenters note major changes since then (printing/formatting “writergate”, build system, C interop).
  • Several say examples don’t compile on current Zig; consider it at least ~2 years out of date and “very brief” / too simplistic.
  • Multiple people assert it appears AI-generated and therefore unreliable for learning.

Alternative Zig Learning Resources

  • Frequently recommended:
    • Learning Zig (online book), zig.guide.
    • “Introduction to Zig” (project-based online book).
    • Ziglings (interactive exercises; kept up to the latest release; listed on the official Zig site).
    • Official “Why Zig over Rust/D/C++” page.
    • TigerStyle docs from the Tigerbeetle project as a style/best-practices reference.
    • zigdoc and ziglint plus zig fmt for tooling.

Versioning, Stability, and Tooling

  • Pain point: many Zig projects need a specific compiler version; mismatch with local install or between projects is common.
  • Some call this a “major miss,” others say: it’s pre-1.0, instability is expected and accepted by early users.
  • Comparisons to Rust editions, which allow mixing versions; Zig lacks an equivalent.
  • Suggested mitigations: use version managers/isolated environments (mise, Nix, Docker, etc.).

Whether to Learn Zig (vs. Rust/C/C++)

  • Pro-Zig views:
    • Simpler, closer to C (with some Go-like aspects), easier to understand what the machine does.
    • Good fit for high-level developers occasionally dropping to low-level for performance, due to simple syntax and strong toolchain/cross-compiling.
    • Even if it stays niche, knowledge transfers to general systems understanding.
  • Skeptical views:
    • Rust already fills most of the “modern systems” space, with memory safety and strong industry push.
    • Zig is unsafe, unstable, and may remain niche or follow D’s trajectory; C remains the dominant systems language.
    • For deep systems learning, C (and maybe Rust) is preferred; Zig may feel redundant if you already know C well.
    • Some question treating languages as long-term “investments”; others counter that fluency and keeping up with changes are real costs.

Language Design, Syntax, and Concurrency

  • Some dislike .{}
    • Viewed as noisy when used for arguments and default/optional/variadic patterns.
    • Complaints about having to explicitly discard unused return values.
  • Defenders explain:
    • .{} is inferred struct/tuple construction, often less verbose than naming types.
    • Explicit discard is intentional for catching bugs.
  • Concurrency model:
    • Said to be broadly Go-like: queues instead of channels, async/concurrent APIs instead of go keyword, and a select-like construct.
    • Some lament the lack of coverage in the example material.

AI, Ecosystem, and LLM Support

  • There’s tension around AI:
    • Some criticize Zig’s project-level bans on AI-generated contributions and the “anti-AI” crowd as harmful to adoption.
    • Others clarify you can freely use AI to write Zig applications; restrictions mainly apply to Zig’s own codebase.
  • For better Zig-in-LLM support, suggestions include:
    • Integrating with the Zig language server, local docs, and source as canonical references.
    • Building MCP servers/tools to query docs, linters, formatters (zig fmt), and real-world codebases.
    • Using Zig project histories as benchmarks to grade AI-generated PRs.
  • A few express that AI has made Rust more approachable and fear Zig may miss that wave; others argue you can always just ask an LLM directly instead of consuming AI-generated tutorials.

Adoption, Community, and Motivation

  • Perceptions vary:
    • Some see Zig as promising due to toolchain, C interop, and a strong standard library.
    • Others predict it will remain niche, with limited corporate adoption and a community sometimes described as hostile or elitist.
  • Usage anecdotes:
    • Hobbyists and non–C/C++ professionals using Zig for drivers, FFI layers, and side projects.
  • One commenter questions whether Zig still “matters” in the AI era and reports loss of motivation to keep learning it.