I have converted my apps to open data.
Users own .app/ and data/. They can use data/ however they want, or interact directly with .app/ through a CLI. As long as they follow the open convention, the choice is theirs.
Why invest in a pixel-perfect UX and a fast backend if you give everything away?
Because intelligence and logic are becoming a disposable layer. For builders and for users.
Builders should be able to rapidly delete, rebuild, and upgrade. Users should be able to do the same.
The software stack is evolving so quickly that the durability of any single app is uncertain. Locked formats will soon become a liability.
The logic layer is now genuinely disposable. Code agents improve constantly. Frameworks rise and fall. If your data is locked in, you slow yourself and your users down. That is not a strategy.
Users will see you as next in line for replacement if you hold them back while everything else accelerates.
There must be a better way.
My view? Embrace it. Open data is the new open source.
Keep the data model in plain text. Build intelligence on top of it. The smart features and the visual polish live in a disposable layer above the data.
Then earn user trust repeatedly. Sell convenience, not lock-in.
Users (and eventually Agents) decide whether to replace you, rebuild you, or extend you with their own agents.
Your application becomes one of many possible view and logic layers over the data. Your job is to make it the best one.
I built a skill to implement these principles, I am actively iterating as I test out different designs but this is what I have so far: open-data on GitHub.