[moron4hire]

>> PostgreSQL supports DROP SCHEMA CASCADE, which drops a schema and all the database objects inside it. This is very, very important for a robust analytics delivery methodology, where tear-down-and-rebuild is the underlying principle of repeatable, auditable, collaborative analytics work.

Actually, if you drop the constraints first, you don't have to worry about the table-drop order. I do that on both Postgres and SQL Server, because I use a tool that generates the specific scripts for me.

I don't think manually writing DDL scripts is any way to manage an RDBMS in the modern era, especially if you care about "repeatability and auditability" like he claims.

Queries, yes, but schema constructs, no. SQL is just a terrible language for it. I'm not saying you have to use a full ORM, but every SQL engine I've ever encountered will let you do a lot of nutty things and won't complain about it, or won't complain until you actually try to query it.