MkDocs/Material/mkdocstrings end of support

[llucax]

What happened?

The last mkdocs-material release is announcing:

Material for MkDocs is in maintenance mode

Going forward, the Material for MkDocs team focuses on Zensical, a next-gen static site generator built from first principles. We will provide critical bug fixes and security updates for Material for MkDocs until November 2026.

Read the full announcement on our blog

Also mkdocstrings-python announced entering in maintenance mode:

• ⚠️ mkdocstrings-python is in maintenance mode mkdocstrings/python#323

What did you expect instead?

To be f*cking developed forever 💔

Affected version(s)

No response

Affected part(s)

Build script, CI, dependencies, etc. (part:tooling), Tools to configure mkdocs (part:mkdocs)

Extra information

I'm pretty sure Zensical is/will be awesome (is not only from the mkdocs-material guy, but it looks they also hired the creator of mkdocstrings). Also for the rust crowd:

Zensical is written in Rust and Python

(of course it is 😆)

But I have the feeling that, unless they are really good to provide tools to migrate people and still somehow support mkdocs extensions natively, this will be a real pain, because we heavily rely on many mkdocs extensions, not only material.

We'll see if there are any forks that keep maintaining their mkdocs stuff.

[llucax]

OK, maybe there is no way out:

Coming from MkDocs?

If your organization builds its documentation with Material for MkDocs, we strongly recommend to start thinking about migration early. MkDocs 1.x has been unmaintained for over 18 months, and the upcoming MkDocs 2.0 will introduce breaking changes for your documentation projects.

The time is now.

The blog post about MkDocs 2.0 offers more details, and they are really worrying:

• MkDocs 2.0 won't have a plugin system – MkDocs 2.0 is being rewritten with a focus on theming, deliberately removing plugin support to simplify the codebase. We believe the plugin ecosystem is one of the cornerstones of MkDocs' success and widespread adoption, and its removal will affect workflows and customizations that teams have built over time.

• MkDocs 2.0 has a closed contribution model – MkDocs 2.0 moves to a "maintainer-driven issues" model, where community members are asked not to open issues or pull requests, and "feature requests are generally not accepted". It's unclear how users should report bugs or seek fixes when they encounter problems.

• MkDocs 2.0 is currently unlicensed – MkDocs 2.0 doesn't specify a license, which has implications for how it can be used and contributed to by the community. It's unclear what the rationale behind this decision is, but it should be critically evaluated by teams and organizations that rely on MkDocs for their documentation projects.

• MkDocs 2.0 has a new configuration format – MkDocs 2.0 uses TOML for configuration, which is incompatible with the YAML format used in MkDocs 1.x. As a result, existing mkdocs.yml files will currently not work with MkDocs 2.0. There is no migration path for existing projects.

Zensical offer some paid support migrating (https://zensical.org/spark):

Join Zensical Spark, and get hands‑on migration support from the team that built and maintained Material for MkDocs for the past 10 years, helping you to make the transition to Zensical seamless. You're investing in the continued growth and sustainability of Zensical, allowing us to actively develop it, keep it reliable, and Open Source in the long term.

[llucax]

They are at least trying:

Zensical is designed to be a drop-in replacement for MkDocs 1.x, with the goal of building your existing projects without any changes. We're not fully there yet – supporting the most popular MkDocs plugins in the ecosystem is an ambitious undertaking – but it's our current focus, and we're fully committed to getting there.

I guess for a few months this is a wait and see how things develop. Maybe we can try to do some early tests to see how far are we from being able to generate our docs using Zensical and provide early feedback. I guess this will increase the chances that plugins we need are supported sooner than later.