That second trick is neat, I'll need to remember that. I also really wish I had known about it back in 2015, when we had to massage an inconveniently big (at th

That second trick is neat, I'll need to remember that. I also really wish I had known about it back in 2015, when we had to massage an inconveniently big (at the time) PG database and started to carve it out into smaller, purpose-specific instances. Being able to verify that an index was either useless or inefficient without jumping through hoops would have saved quite a lot of time.