I'm very much just a basic user of Docker and Compose, but whenever I run into issues with Docker I see how v3 or the new `docker compose` CLI does things differently, and I'm just wondering, what was wrong with the current way of doing things?
`docker-compose -f` used to work fine, I could specify a file but `docker compose -f` doesn't. On top of that, as soon as I use a none standard filename (for e.g. a development workflow) Docker Desktop isn't able to stop the entire Compose project (I have to go through each container individually) nor restart it later on.
I was also reading about `healthcheck` as well as `depends_on`. It seems that now they want to move to using tools such as [`dockerize`](https://github.com/jwilder/dockerize) to handle these basic things, and I don't really get why something that directly has to do with the fact that Docker Compose is managing the infrastructure should be moved to near application-level logic. It adds more boilerplate in a decentralized place, without actually decoupling since there's more than enough to configure in the Compose file anyway.