I don't think it's a good idea for Rust to choose a particular kind of concurrency primitive. Rust is pretty broad in its use cases, and choosing a primitive may end up restricting other fields. Remember that Rust used to have a libuv based green threading model, and it was removed for precisely this reason -- it affected some use cases in major ways.
Rather, it is important for subcommunities to rally around particular sets of libraries. E.g. the web and I/O related libraries seem to all be going in the direction of integrating with tokio. This means a web programmer can use tokio and it will work smoothly. OTOH, a different subcommunity that uses a different set of libs may use stuff built on rayon and crossbeam.
> Hmm. I guess the community will eventually decide on one particular primitive, hopefully soon.
Ultimately you'll get libraries or frameworks that are not compatible with each other because they use different concurrency models. That's a mistake Rust will pay in the future.
You get that anyways. The question is how popular something like tokio will get. Since its blessed by the core team and the language designers, I think it will get popular, and therefore many will use it. But there will inevitably be someone who doesn't like it, or for whom it doesn't fit, and they will use something else. But that happens no matter if tokio is in the stdlib or not.
That's obviously a threat. But so far this is not what's happening: since tokio and future-rs were announced we've seen most library authors ditching their old stuff to use tokio instead.
Not really. People in the Rust community have shown a great deal of willingness to standardize on a handful of traits, so things tend to be intercompatible. That might change, I guess. In any case, there's no point in using an async I/O model for the majority of use cases many of the people using Rust are interested in, so people standardizing on something else for those usecases isn't a problem.
I hope the same, but i don't have great hopes. For a community to adopt a library as fundamental as this one would be, it would need to be of such a high level that i think only the language designer can create it ( or other really big names in the field).
If you think of something like OTP, i don't see any equivalent anywhere else. Even akka isn't unanimously used as the foundation for server side dev on the jvm.
Two of the three Tokio core team members are on the core team of Rust, and one of them is on the language design team.
EDIT: to be clear, that means one is on both. The third member is very well-known as well; having built the async primitives that this is built on, as well as initially implementing Cargo. Bona fides aren't an issue here.
