Summary

Armin Ronacher critiques Codeberg's new terms excluding projects 'mostly' written by generative AI, acknowledging the association's democratic right to make the decision while arguing it's a poor one. He distinguishes between democratic legitimacy and good governance, noting that infrastructure must above all be predictable and neutral. The vague 'mostly' threshold makes the rule unenforceable and delegates policy to moderators and community norms. He argues a sharper rule targeting spam or resource abuse would be more effective. Ultimately he sees the decision as pushing Codeberg toward niche community platform rather than broad GitHub alternative, at a moment when the open source world needs to engage with AI rather than divide over it.

Key Insight

Codeberg's vaguely worded AI policy is unenforceable in practice and risks trading its potential as a broad, neutral GitHub alternative for a smaller, politically homogeneous community — exactly when open source needs to engage with AI's future rather than divide over it.

Spicy Quotes (click to share)

• 8

  democracy is a way of making a decision, not a guarantee that the decision is inclusive, wise, or even good for the people already depending on it

• 7

  A democratic provider without a clear constitution can be worse at those things than a corporation.

• 6

  The line is open to interpretation precisely where it needs to be enforceable.

• 4

  I'm currently assuming the community around it draws a much harsher social boundary, making projects and maintainers unwelcome even when they technically comply.

• 4

  The current middle ground delegates too much of the policy to moderators and community norms.

• 6

  LLMs if done and used well, should be welcome to all of us. They could be used to reclaim control and power, away from large corporations and institutions.

• 7

  I wish Codeberg were more forward-looking here: willing to host the Open Source software of tomorrow, not only software made in the ways its community approves of today.

Tone

critical, pragmatic, measured