Comment by armchairhacker
4 years ago
Usually I think trying to make your own language for anything more than a DSL is a bad idea. But this project looks very impressive.
I've never heard of Imba before this post and haven't seen any real-world projects (besides Scrimba). If this is as good as it looks we'll probably get some soon.
Reminds me of Elm which took a similar route and seems to be successful.
As someone else said, one thing you probably want is static typing. Also make sure your language works well on larger projects. Everyone thinks they're making a simple website, but then it scales.
> Usually I think trying to make your own language for anything more than a DSL is a bad idea.
Depends what "your own" means. If it's a one-off language just for one specific project, then you're right. But this one is much more than that. Surely you don't think people should stop making new languages entirely?
No i mean that a single person or small group can’t make a general-purpose language that would beat TS, Rust, Java, Swift, Go, etc.
All these languages are supported by huge groups. They have hyper-optimized compilers or JIT, IDE integration, libraries, and adoption.
If you can form a huge group, then you can beat these languages. Or if you target a niche. Or if you can use existing tooling (transpiling, LSP, effective existing libraries), and spend a lot of time and effort. Imba seems to be taking the third approach.