PostgreSQL 19 lost 8 to 10 major features after beta began — and LLM code review is partly why
The reverts will continue until morale improves
Peter Eisentraut, a PostgreSQL committer, reflects on why an unusually large number of features were reverted from PostgreSQL 19 after beta started. He credits LLM-assisted code review with surfacing long tails of edge-case defects faster than ever, notes that senior developers were tied up fixing security issues during the Vulnpocalypse, and argues the community should adopt experimental feature status and more robust internal interfaces.
I like to say, in PostgreSQL, "everything works with everything". This is part of the appeal, that you can construct cases like this and expect them to work. I don't want to give up on that at all. But it does impose a significant penalty on feature development that everyone needs to be aware of.