I thought for quite a few days on the Jochen’s study for the OSM data model, and I’ve no idea on whether it’s good or bad or how to improve it. The process feels pretty straightforward to me.

But that’s the issue. It’s not impossible. To fix OSM data issues, changing the data model is not important. Frankly, I don’t have anything to add to my 2019 talk at Heidelberg. It was a reply to Jochen’s and Andy’s musings on API 0.7 back then, and it is still now.