> aren't CTEs optimization fences in postgres? Apparently "not all the time with PG12". https://www.depesz.com/2019/02/19/waiting-for-postgresql-12-allow-user

> aren't CTEs optimization fences in postgres? Apparently "not all the time with PG12". https://www.depesz.com/2019/02/19/waiting-for-postgresql-12-allow-user-control-of-cte-materialization-and-change-the-default-behavior/ https://www.depesz.com/2019/02/19/waiting-for-postgresql-12-...