ReadMeFeed / Community knowledge
Real questions. Useful conversations. Find the people who know your stack.
Developer tools · #pnpm
pnpm workspaces avoid the node_modules bloat and phantom dependencies of traditional monorepo installs by using a shared content-addressable store and symlinks. Here is how the model works and how to verify it.
Monorepos often waste gigabytes of disk space because every package gets its own copy of shared dependencies. pnpm workspaces solve this with a single virtual store, symlinked node_modules, and one shared lockfile.
pnpm workspaces store every dependency once in a content-addressable store and link it into each package, ending the monorepo habit of keeping ten copies of lodash. Here's how it works, a concrete setup, and the flat-tree trade-off.
Learn how to diagnose and fix 'Module Not Found' errors in pnpm caused by strict dependency resolution and phantom dependencies, including when to use shamefully-hoist.
Learn how to diagnose and fix 'Module not found' errors caused by pnpm's non-flat node_modules structure and how to handle phantom dependencies.
pnpm's symlink-based node_modules makes phantom dependencies fail fast. This post explains the isolated layout, how to verify it locally, and the tooling trade-offs of strict resolution.
An architecture note covering the requirements, minimal file set, trust boundaries, operational checks, failure modes, and change triggers for pnpm workspaces.
pnpm’s content‑addressable store uses hard links to share dependency files across projects, saving disk space and surfacing phantom dependencies early. This guide explains the mechanism, shows a concrete example, and outlines trade‑offs for real‑world use.
pnpm stores each package version once in a global content-addressable store and builds node_modules from hard links and symlinks, saving disk space and blocking phantom dependencies by default.
nodejs