uv-build Draws Gentoo Portability Backlash
A Gentoo developer urges Python maintainers to reconsider uv-build for pure-Python packages, arguing its Rust dependency, frequent releases, and 1.2–1.5 MiB wheels impose disproportionate costs on downstream packagers and uncommon architectures. The critique targets uv-build specifically, not Astral’s broader uv toolchain. [Source](https://blogs.gentoo.org/mgorny/2026/10/12/a-plea-not-to-use-uv-build/)
uv-build is a sensible convenience for uv-native projects, but making it the default can shift small build-time benefits into real portability and maintenance costs for everyone downstream.
- –PEP 517 makes build backends interchangeable, so uv users can still build projects using flit-core, hatchling, or setuptools.
- –The backend currently targets pure-Python projects, where its performance advantage is unlikely to matter compared with dependency and compilation overhead.
- –Gentoo’s portability concern is especially relevant to source distributions, alternative architectures, and systems where Rust tooling is difficult to bootstrap.
- –Astral documents uv-build as optimized for portability and small binary size, making this dispute a useful test of whether those claims hold across distribution ecosystems. [Official documentation](https://docs.astral.sh/uv/configuration/build-backend/)
- –Maintainers should choose uv-build for concrete project needs rather than treating alignment with uv as sufficient justification.
DISCOVERED
1h ago
2026-10-13
PUBLISHED
3h ago
2026-10-13
RELEVANCE
AUTHOR
Ravencentric