Kyle Gray

smallint maps to a int16, so I'm not sure why changing the array to int32 was necessary. Will take a look

I think this is a great idea. When the 1.18 beta was released, I [asked for ideas](https://github.com/kyleconroy/sqlc/discussions/1342) on how sqlc should use them. Allowing people to experiment with nullable wrappers...

@krhubert Wow, that's one... inventive workaround. It's like SQL injection but for Go code generation. I think `imports` vs `import` works, but need to make sure the code generated constructs...

The configuration options don't need to be changed. Version 2 of the configuration files supports per-package renames and [global renames](https://docs.sqlc.dev/en/stable/reference/config.html#global-overrides). The bug is that we aren't pulling in the rename...

This has been fixed https://play.sqlc.dev/p/8b270e64462166a51326f9c26712b83cc39b5e69c67617b6d14185604ab33e07

Whoops, I checked the first case instead of the second case. Reopened

Can you please provide a few sample queries that exhibit the issue?

I think @eullerpereira94 summed it up better than I could.

> I feel like you don't need to tell sqlc about the exact schema because it doesn't ultimately manage the schema. sqlc expects to have an accurate view of your...

All four of these are CockroachDB extensions or features. Primary key ordering isn't a PostgreSQL feature, at least not as of PostgreSQL.