Timo Stollenwerk: How a contributor-led open-source platform decides what it becomes — Plone roadmap and AI, in public

Choosing the right system is hard. There are no easy solutions. Where does the platform go after you have signed the deal? There is no free lunch and no risk free project. It is about mitigating and lowering the risks.

Biggest risk: single-vendor lock.

Then there are geopolitical and jurisdictional risk. The Brazilian government is using Plone a lot, so you have a choice. Own your digital future. That is a theme of this conference.

You can choose between closed and open source solutions. Everyone here will agree that open source is the better choice. Careful: there is a difference between open source versus free and open source. Vendor open source versus real open source: some solutions are driven by a single vendor, in other cases it is multiple vendors. Then there are vendors that contribute versus freeriders.

Some CMS systems have a paid add-on system, others a free ecosystem. Do you build or do you buy? Do you want your skills in-house or are you fine with hiring an external company. This is not black and white. Third option between buy or build: join.

How do open source communities work? How do we make decisions? Plone has a board of directors, we have release managers, teams, a steering circle meeting regularly where teams report and synchronize.

We have a Roadmap Team. How do we get our roadmap? We have the PLIP process: PLone Improvement Proposal. A PLIP is just an issue in a ticket system. Anyone can submit a PLIP. Create a new issue in the issue tracker and select the PLIP category. When working on it, you do need to sign a contributor agreement, for legal requirements.

Plone is a do-ocracy. If you make a proposal, you should also have a willingness to work on it. But you can get others involved, and it is also open for non-coders.

"The ladder": before working on a PLIP, start with small contributions, so you get to know how Plone and the community work.

How do you get from a PLIP to a roadmap? Let's focus for this day on an AI roadmap. There are different opinions here.

Look at PLIP 4213: Plone Model Context Protocol Server. This gets discussed. Should this be in core or an add-on? Should it be in JavaScript or Python or something else? Sometimes we would say: first make it work, maybe others make a different variant, try out alternatives, see what works in practice.

So we have a do-ocracy with a transparent process. We have a few formal quality gates. The Framework Team can make a final decision if needed when the discussion is heated. In reality this is not so much needed.

The AI roadmap is nothing more than a project on GitHub. We take those PLIPs, and at the Plone Conference one of the release managers presents the plans. Around or after the conference the roadmap team talks with each other. Teams like Blicca, Volto, and Marketing can get involved to get perspective.

Plone has 25 years of experience and track record and counting. Plone and the Plone community won't go away. Plone is multi-vendor and real open source. We have the Plone Foundation, really important, a solid legal foundation. Younger projects often have lots of enthousiasm, but not these legal guard rails.

If you build: we have a community you can join. If you choose to buy: we have a community recognition program.

Open source communities have to learn and adapt. There should be a contract or quality standard for all open source projects. Do they have a legal foundation? How long do they exist? If we want the wider people to adopt open source, we have to make it easier for them to make the right choices. This goes beyond the Plone community.