The main innovation here seems to be compile time checking of that foreign keys are respected but that is a thing that can be added to SQL and there is at least

The main innovation here seems to be compile time checking of that foreign keys are respected but that is a thing that can be added to SQL and there is at least one proposal for doing so. So I do not really see anything fundamentally different from SQL. https://keyjoin.org/ https://keyjoin.org/ Full disclosure: I am one of the co-authors of this paper and an associated patch implenting it in PostgreSQL that we have proposed. I am happy to see more people than us think this is useful.