Git repositories seem like an elegant solution for package registry data. Pull requests for governance, version history for free, distributed by design. But as registries grow, the cracks appear.

Lobsters is a community-driven platform for sharing and discussing links to articles, tutorials, and projects related to technology and programming. Readers can learn about a wide range of topics, from software development and system administration to cybersecurity and artificial intelligence. With user submissions, comments, and voting, Lobsters provides a platform for collaborative learning and knowledge sharing among technology enthusiasts.

Lobsters

Git repositories initially seem ideal for package registries with built-in version control and distributed architecture, but major package managers like Cargo, Homebrew, CocoaPods, and Go modules all eventually abandoned git-based indexes due to performance problems. As registries grow, git's filesystem limitations cause slow clones, expensive delta resolution, and CI bottlenecks. The pattern is consistent: start with git, hit scaling issues around directory limits and cross-platform compatibility, implement workarounds like sharding, then migrate to HTTP-based protocols or CDNs. Git lacks database features like constraints, indexes, and efficient point queries that package registries actually need.

Package managers keep using git as a database, it never works out

<p>Makes sense. I wonder if we’ll see a new wave of preference towards vendoring all dependencies.</p>
