My advice would be don't bite off so much at once where you'll risk getting discouraged. Part of the reason you may be feeling overwhelmed is that it sounds like you're combining three projects: learning a new programming language, learning network programming, and writing an application server in a new (to you) domain. Any one of them is potentially enough to keep you quite busy.
Why not instead start with a language you already know, and figure out how you'd sketch out a standalone game engine (forget about networking for now) in that. Then once you think you've got the basic game engine (architecture, at least) down, then tackle turning it into a network server (again, still using a language you already know.) Finally, port the thing over to a language you want to learn (Elixir/Scala/whatever) and you'll have an implementation you understand well to compare it against. Of course you can rearrange the sequence... but that's the basic idea.
I think you're definitely right, I'll try to give it a try using javascript, and learn the basics of network server along the way. The reason I wanna do it in elixir is that I wanna learn another language, right now javascript is the only one I can say that I know kind of well, all my side projects are in js, so I guess I'm a bit bored of it.
Why not instead start with a language you already know, and figure out how you'd sketch out a standalone game engine (forget about networking for now) in that. Then once you think you've got the basic game engine (architecture, at least) down, then tackle turning it into a network server (again, still using a language you already know.) Finally, port the thing over to a language you want to learn (Elixir/Scala/whatever) and you'll have an implementation you understand well to compare it against. Of course you can rearrange the sequence... but that's the basic idea.