November 2025 Summaries

The discussion revolves around optimizing the project search feature in Zed, focusing on improving its performance without sacrificing long-term maintainability. Despite the team's dedication to performance, they avoid extreme measures like delving into compiler-generated machine code or hand-writing SIMD. The solution involved restructuring the task hierarchy to prioritize confirming matches in files over scanning new files for potential matches, thus enhancing responsiveness. While this approach improved responsiveness and throughput, it highlighted ongoing challenges, such as latency in remote projects and UI rendering issues when handling numerous matches. The team continues to strive for parity with ripgrep, a popular search tool, while acknowledging the complexity of project search and the need for further improvements.

Zed Industries has developed a collaborative text editor called Zed, designed to enhance teamwork among distributed teams by offering a seamless and integrated collaboration experience reminiscent of pair programming at Pivotal Labs. Unlike traditional editors that require cumbersome setups or extensions for collaboration, Zed utilizes CRDTs to ensure conflict-free and consistent merging of edits in real-time, providing a low-latency experience regardless of team size or geographic location. The platform includes features like built-in audio, screensharing, and a flexible channel system for organizing workspaces, which accommodates company-wide discussions, project-specific collaborations, and personal focus spaces. Through its GPU-accelerated UI rendering framework written in Rust, Zed maintains high performance even as more collaborators join. The team leverages Zed internally to build the product itself, continually refining it to align with their vision of creating the ultimate multiplayer software development tool. Although collaboration features are still in their alpha stage, they are available for free, with ongoing development aimed at integrating AI agents and continuous conversation capabilities to further enhance the collaborative experience.

The Agent Client Protocol (ACP) ecosystem is expanding rapidly with new agents like Auggie from Augment Code and clients such as the JetBrains family of IDEs. Zed has introduced Agent Server Extensions, enabling users to easily install and use agents like Augment, OpenCode, and Stakpak with a single click. These extensions add a menu item in Zed to start a new thread with an agent and automatically download the necessary implementation, though authentication is typically required. Developers can create ACP-compatible agents and distribute them through Zed by following simple steps to publish their own extensions. The community and Zed team are actively supporting the growth of this ecosystem, with related content and opportunities available for those interested in software development.

Addy Osmani, from Google's Chrome Developer Experience team, discusses the "70% problem" in AI coding during the Agentic Engineering Sessions. He highlights that while AI tools can generate up to 70% of code efficiently, the remaining 30%—involving edge cases, security, and production integration—remains challenging, often negating the initial productivity gains. Despite increased adoption, trust in AI-generated code is declining, with many developers reporting low confidence in its reliability. Osmani emphasizes the importance of human oversight, urging engineers to understand and review AI-generated code to ensure security and maintainability. He also notes that productivity gains touted by some companies are often context-dependent, particularly in greenfield projects lacking technical debt. Osmani advises junior developers to use AI as a learning tool while maintaining curiosity and a commitment to understanding the code they work with.