> this seems like such a cop out. its like saying "oh failure is a given, so don't even try to succeed".

It is a pragmatic cop-out yes. I found it is easier to make progress by assuming that you'll get to Turing Complete rather than investing in the time to avoid it. I found that the downsides of Turing Completeness are usually overhyped or primarily theoretical.

> so just because Rust (and others) failed, doesn't mean its impossible.

It definitely is not impossible. But I don't know if it is worth it.

In the end I want fast compile times, type safety, easy to maintain code and good error messages. I do not care if the type system is Turing Complete.

> In the end I want fast compile times, type safety, easy to maintain code and good error messages. I do not care if the type system is Turing Complete.

I think the problem is some languages with Rust go too far with generics, which probably triggers the turing complete. for example, this is valid Rust code:

    let mut handles = Vec::new();
    for i in 0..10 {
        let handle = do_work(i);
        handles.push(handle);
    }

but you have to follow the code all the way to "do_work" before you ever find the type of anything. Go does not allow this. you need to either declare a concrete type:

    var handles []int

or a explicit generic type:

    type slice[T any] []T
    var handles slice[int]

I think Rust is lose lose, because the underlying type implementation is Turing complete, and I would argue the code is actually less readable because of the overuse of type inference.

> I think Rust is lose lose, because the underlying type implementation is Turing complete

I am not a Rust coder so I can't really comment.

Currently, I do TypeScript, which also has a Turing complete type system, and I love it. Of course all things in moderation. Even though I could make a completely obtuse type design for projects, I try to not write code that others can not understand.

I don't think most people care at all that their type system is turing complete. It basically never comes up.

Agreed. Adding to this point, why should I care? Am I missing anything obvious on why you don't want it?

Not really. In theory it means your compiler might never terminate... that does not usually happen.