Rust Glancer: A Low-Memory Alternative Rust LSP
Original: Rust Glancer: Rust LSP using 100x less RAM
Why This Matters
Demonstrates a viable low-resource alternative to rust-analyzer, expanding Rust tooling accessibility on memory-constrained hardware.
Developer releases Rust Glancer, an alternative LSP for Rust built over 4 months, targeting under 100MB RAM usage. It supports persistent indexing across editor restarts, full type inference, a trait solver, and core LSP features like goto definition, hover, and completions.
Rust Glancer is a new Language Server Protocol (LSP) implementation for Rust, developed over four months with a primary focus on low memory usage. The project targets under 100MB RAM for typical projects and enables persistent indexing — meaning that after an editor restart, previously indexed projects do not need to be re-indexed from scratch.
Benchmarks show competitive indexing speeds: on a MacBook Pro M4 Max (36GB, 2025), Rust Glancer reached base indexing in 5 seconds vs. rust-analyzer's 6 seconds, with full indexing in 8 vs. 13 seconds. On an older MacBook Pro M1 (8GB, 2020), Rust Glancer achieved 6 seconds (base) and 9 seconds (full), compared to 7 and 14 seconds for rust-analyzer.
Architecturally, Rust Glancer diverges from rust-analyzer by avoiding incremental analysis. Instead of using salsa (an in-memory incremental query database) and rowan (a syntax tree with partial invalidation), it produces a frozen analysis result invalidated on save. This allows analysis data to be offloaded to the filesystem and loaded into memory only when needed, directly reducing RAM footprint.
The project already includes type inference, a chalk-based trait solver, and standard LSP actions. A VS Code extension is available now. The developer acknowledges missing functionality and known bugs given the short development timeline.