Hello Said,
Responding to the questions:
1. I'm convinced we should aim for the newest version. It was released 3 months ago, and if editor support is not there yet, it should definitely arrive by the time the project is done.
2. It's an interesting feature - I don't think it's a basic requirement, but definitely a "nice to have".
3. The first goal should be with regular elixir code. That said, while ecto or phoenix contain their fair share of macros, they are "just" regular elixir applications, so there shouldn't be that much to do to support them.
4. Yes, we can definitely reuse some parts of alchemist-server. I already talked with tonini - the author of alchemist-server and we was very supportive of this project.