Best of Zed — 2025

  1. 1
    Article
    Avatar of zedZed·24w

    How We Rebuilt Settings in Zed — Zed's Blog

    The Zed team rebuilt their settings editor from scratch, moving from a distributed, macro-based architecture to a centralized, strongly-typed model. The refactoring consolidated scattered settings definitions into UserSettings and ProjectSettings types, treating configuration files as the organizing principle rather than implementation details. They enhanced their GPUI framework with automated focus handling through tab groups and local state management similar to React's useState. The project revealed architectural weaknesses (including an auto-update bug) but resulted in a cleaner foundation for future features.

  2. 2
    Article
    Avatar of zedZed·32w

    Windows When? Windows Now — Zed's Blog

    Zed code editor launches stable Windows support with native DirectX 11 rendering and DirectWrite text integration. The release includes full WSL and SSH remoting capabilities, allowing developers to edit files on Linux systems directly from Windows. All Zed extensions work without modification through WebAssembly Components and WASI sandboxing. AI features including edit predictions and ACP-powered agents are fully supported on Windows. The team maintains dedicated Windows engineers and will ship weekly updates matching their Mac and Linux release cadence.

  3. 3
    Article
    Avatar of zedZed·47w

    Why I'm Dialing Back My LLM Usage — Zed's Blog

    A seasoned software engineer with 15 years of experience shares his honest reflection on using LLMs in production code. After initially embracing AI tools with enthusiasm, he encountered significant challenges including poor code quality, cascading bugs, and productivity illusions. He advocates for a more measured approach where developers maintain architectural control and use LLMs only for small, well-scoped tasks like refactoring rather than large features. The key insight is treating LLMs as assistants rather than replacements, emphasizing that senior developers should trust their skills and use AI to leverage existing knowledge rather than offload critical decision-making.

  4. 4
    Article
    Avatar of zedZed·25w

    Zed Has Rainbow Brackets — Zed's Blog

    Zed code editor introduces rainbow brackets, a highly-requested feature that colors nested brackets at different levels for improved code readability. The implementation uses a chunk-based approach with tree-sitter queries rather than maintaining full syntax trees in memory. Brackets are colored in 50-row chunks that invalidate on buffer changes but only re-query when visible, avoiding the complexity of global state management. The feature is opt-in per language, reuses theme accent colors, and includes performance optimizations to tree-sitter's query execution.

  5. 5
    Article
    Avatar of zedZed·36w

    Hired Through GitHub: Part 1 — Zed's Blog

    Zed Industries shares stories of developers who joined their team through open source contributions rather than traditional hiring. The post highlights two contributors: Junkui Zhang, who became their top external contributor by implementing Windows support features, and Anthony Eid, who worked on the highly-requested debugger feature. Both developers started as community contributors, engaged through GitHub and Discord, demonstrated consistent code quality, and eventually transitioned to full-time team members after building relationships through pair programming sessions.

  6. 6
    Article
    Avatar of zedZed·41w

    Why LLMs Can't Really Build Software — Zed's Blog

    Large Language Models excel at writing code but struggle with the iterative mental modeling that defines effective software engineering. While LLMs can generate code and update it when given specific problems, they cannot maintain clear mental models of requirements versus implementation, leading to confusion when tests fail or debugging is needed. Current models suffer from context omission, recency bias, and hallucination issues that prevent them from understanding complex software systems. For non-trivial projects, human engineers must remain in control, using LLMs as tools while maintaining responsibility for requirements clarity and code verification.

See all Zed archives