Collected information from members of Scientific Python for the Anaconda.org User Research study interview @matthewfeickert has scheduled on 2025-05-07.
Service so far has been quite useful, especially for providing infrastructure for the whole development community to be able to test against collectively.
Specific examples:
Having an obvious place to put free-threaded wheels was really nice for some projects depending on NumPy getting off the ground before numpy 2.1 came out
Features
import micropip
await micropip.install(["numpy"], pre=True, index_urls="https://pypi.anaconda.org/scientific-python-nightly-wheels/simple")
import numpy as np
np.__version__ # 2.3.0.dev0+git20250503.00f2733
Security
Collected information from members of Scientific Python for the Anaconda.org User Research study interview @matthewfeickert has scheduled on 2025-05-07.
Service so far has been quite useful, especially for providing infrastructure for the whole development community to be able to test against collectively.
Specific examples:
Features
cibuildwheelto build WASM/Pyodide wheels forscikit-image, push nightlies to Anaconda.org scikit-image/scikit-image#7440 (comment) it appears that this has been resolved in 2025 (Web proxy to work around the lack of CORS headers for nightly wheels uploaded to the Anaconda.org index pyodide/pyodide#4898 (comment)) and there is now at leastscikit-learnsupport (DOC Use nightly WASM wheels for JupyterLite in the dev documentation scikit-learn/scikit-learn#31085).https://anaconda.org/scientific-python-nightly-wheels/groups/<package name>)https://anaconda.org/scientific-python-nightly-wheels/groups/<package name>/settings)https://anaconda.org/scientific-python-nightly-wheels/groups/<package name>/packages)Security