I'm a big Rust fan, so I'm happy Tokio holds up so well. I find it to be one of the simplest and quickest ways to write task based programs.
Gotta wonder whether you can challenge those numbers with c++ though. I suspect you can but with a lot of work, where Rust is pretty effective written naively.
Of course the big issue with these comparisons is it's hard to be an expert in so many different languages. How do you know if what you're doing is the best way in each language?
I find the opposite to be true, Tokio is a fine library but it makes the whole program very complex.
Perhaps it's because I generally don't write code for many short lived workloads, but I find using real threads to be a lot easier to work with (plus it saves time during compilation). Tokio is great at what it does, but I have to wonder how relevant these types of threads are when "fearless concurrency" is one of the token advantages of the language.
I think this article would've been a lot better if it did some actual calculations in the spawned threads, rather than simply waiting. A smart enough compiler would see that the results of the wait calls, which usually don't come with any other side effects, are never used and optimize them out entirely.
Gotta wonder whether you can challenge those numbers with c++ though. I suspect you can but with a lot of work, where Rust is pretty effective written naively.
Of course the big issue with these comparisons is it's hard to be an expert in so many different languages. How do you know if what you're doing is the best way in each language?