54 posts tagged with "conda"

conda and conda-build 26.9.0 are out. Since our June and July update, conda-pypi, constructor, rattler, grayskull, and several other projects have shipped releases too.

The June and July 2026 releases included updates to conda, conda-build, conda-libmamba-solver, conda-pypi, constructor, menuinst, rattler, and more! 🎉 All of these have been released to both defaults and conda-forge channels.

Here is a situation a lot of conda users know: you are working in conda environment, you reach for one more package, and it isn't in any conda channel you have configured. The conda way forward isn't obvious, so you run pip install <some-package>. The package installs successfully, your work continues, and the decision appears harmless. Then, perhaps the following day or several months later, the environment breaks while you are installing, updating, or removing something that seems completely unrelated.

The May 2026 releases span conda, conda-build, conda-libmamba-solver, conda-pypi, and conda-rattler-solver! 🎉 You'll find them on the main and conda-forge channels.

For conda, lockfiles are a first-class workflow now, the CLI is lighter on everyday commands, and there is more under the hood for channel metadata and future dependency specs. Step-by-step opt-in for beta previews—Rattler and conda-pypi—is on New features to try.

Preface​

This post is adapted from a talk I gave at PyConDE & PyData 2026.

In it, I walk through the evolution of Python packaging, from distutils to pip, and later to conda, and how these tools emerged in response to different needs within the Python community.

The ideas in this post are based on my experience as a maintainer in the conda ecosystem and conversations I’ve had with users and contributors over time. A recurring theme in those discussions is the tendency to frame conda and pip as competing tools.

What I’ve found more useful is to view them instead as parts of different ecosystems, shaped by different constraints and priorities. That perspective also helps explain why mixing them can sometimes lead to unexpected behavior in practice.

This post is an attempt to share that perspective in a more structured form.

A vision for conda in 2026 is a package manager that is fast (real-time progress, sharded repodata, lighter shell integration, Rattler integration), trusted (better error messages, smarter confirmation friction, streaming output), and delightful (sensible defaults, helpful suggestions, intentional visuals, accessibility). Beyond the core CLI, the roadmap includes native PyPI wheel support, declarative conda.toml project environments, single-binary installers, and a comprehensive API for IDE and agent integrations.

The March 2026 releases included updates to conda, conda-build, conda-libmamba-solver, conda-content-trust, conda-rattler-solver, conda-standalone, constructor, and conda-pypi! 🎉 All of these have been released to both defaults and conda-forge channels.

The January 2026 releases included updates to conda, conda-build, and conda-pack! 🎉 All of these have been released to both defaults and conda-forge channels.

The Q1 2026 Conda CLI roadmap update highlights faster performance, safer ways to work with PyPI packages, and progress toward more reproducible environments.

Welcome to another quarterly update on what shipped in conda CLI and what we're building next. These posts complement our project board by pulling out the highlights and showing where your feedback matters most.

Modern AI agents like Claude, Cursor, OpenCode, and Zed can fetch web content, run shell commands, and even install packages. But they lack direct access to the rich, structured metadata embedded in conda packages. This information is essential for solving complex packaging problems. conda-meta-mcp provides that missing link.