docs: reinstate baseline adopter counts for growth and graduated stages

[dsterz]

To provide a clear signal of stability to enterprise adopters, I recommend we reinstate the baseline adopter metrics we previously agreed upon.

This directly serves the NeoNephos mission, which states that our projects "are intended for serious use in production and mission critical environments".

By defining strict adopter counts, the TAC can ensure that Graduated projects have proven their real-world reliability, fulfilling our promise to provide high-quality, production-ready building blocks for cloud and edge stacks.

[jakobmoellerdev]

nit: production can be defined extremely differently based on context and size of deployment of a tool. Thus I would heed caution here and we should maybe make it more defined

[dsterz]

This is an excellent point @jakobmoellerdev, i agree that "production" for a small developer CLI tool looks very different than "production" for a massive cloud orchestration platform. Attempting to write a rigid, one-size-fits-all definition of "production" into the governance policy would create unnecessary bureaucracy.

[fmui]

What about this? It adds a bit more precision.

The project must demonstrate to the TAC that it is being used successfully in production, or adopted as a dependency in a released product or open-source project, by at least two independent organisations or projects. The TAC will assess whether the evidence presented is adequate in quality and scope, taking into account the nature of the project. Where specific usage details cannot be disclosed, a brief written attestation from each such organisation or project confirming its use will suffice.

[dsterz]

I also see your good pint @fmui that adding a sentence gives the TAC discretion to judge the "quality and scope" based on the nature of the project and i feel this is the the right approach. Not suprisingly this the same concept that was part of the original June 2024 NeoNephos Lifecycle Policy, which is fair and fine and very transparent to have it discussed and agreed upon in public with everyone while reaching somehow consensus.

[fmui]

See comment above - just with five usages.

[hardl]

Imho not only the term "production" can be a bit fuzzy, but also "successfully" ;)
Also "being used" probably should be explained in more detail, especially if it covers also a non-direct dependency.
E.g. let's assume a product adopted NeoNephos' "Project Foo", which has a dependency to NeoNephos' "Project Bar", does that mean "Project Bar" can claim being used by the Product? Would the AC even be fully met in that example, since there are 2 "production" adopters from 2 different organisations, i.e. the product org and NeoNephos itself (via Project Foo)?

[fmui]

See my comment above.
It's really difficult to cover all possible cases. Therefore, I suggested this sentence: The TAC will assess whether the evidence presented is adequate in quality and scope, taking into account the nature of the project.

[dsterz]

Fmpov this is a great direction, and I appreciate how we are collaboratively shaping this!

Building on your good proposal to give the TAC the flexibility to evaluate 'quality and scope' on a case-by-case basis, I think we have a great opportunity to pair this agility with our core NeoNephos value of transparency.

Since 'production' looks different for every project, allowing the TAC to use its judgment is exactly the right move. To make this work nicely, I propose we add a simple requirement for the TAC to publicly document their reasoning when they approve a project's maturity.

Doing this in both the Growth and Graduated stages achieves two very positive things for our foundation:

1. It builds ultimate trust: Enterprise adopters can read exactly why the TAC deemed a project ready, giving them the confidence to adopt it.

2. It mentors future projects: By publishing these assessments, we create a public track record of what 'good' looks like, which will greatly help new projects understand what they need to aim for.

I suggest we update the PR with the following text for both stages:

Growth Stage: The project must document that it is being used successfully in production by at least two independent organisations. The TAC will assess whether the evidence presented is adequate in quality and scope, taking into account the nature of the project. To ensure transparency and guide future projects, the TAC will publish a public summary of its assessment and reasoning prior to the final approval vote.

Graduated Stage: The project must verify production use by at least 5 independent organisations, subject to formal TAC due diligence. The TAC will assess whether the evidence presented is adequate in quality and scope, taking into account the nature of the project. To ensure transparency and build ecosystem trust, the TAC will publish a formal report detailing its assessment and reasoning for public review prior to the final approval vote.

What do you think of this hybrid approach?

It gives the TAC the flexibility they need while keeping our governance beautifully transparent.

[ScheererJ]

I am surprised that after the comments by two reviewers that production might not be a good term it still shows up in the proposal. Let's get rid of it.
From my point of view, a more realistic approach would be that each project has an adopters page and the users of the project can open PRs to add themself to the list. This is doable and provides some general information. After all, it is hard to get real evidence that a company is using a project, but if at least an employee of a company takes the time to open a PR it seems like some invest into the project.
WDYT?

[dsterz]

I can see your confusion @ScheererJ , I personally read it like we need to be more clear about what it means and how it is defined not that's an issue to have it in generel.
I actually would prefer to keep it rather than dropping it.
"production (ready)" is a well known term and strong positive signal towards a proven use, quality etc.

However i imply we are targeting a production metric thorugh industry adoption like it has been defined in the very first lifecycle policy before removal or who it is in the CNCF, which does not mean we have to keep it, thats why i like that we take a moment to think about the implications for the foundation and its endusers (here Donators + Adopters)

The other route we could take is like e.g. the Eclipse Foundation where the production metric is rather an evaluation of the presence of an adopter ecosystem, graduating there to a mature phase depends primary on proving an active adopter community.

These are my thoughts on this.