This week in Rust had a strange article that anonymously kept praising “this framework.” Only the source code twice references hyperlane. Having never heard of it, I DuckDuckGoed it and various articles also point to dev.to and are all 404. Even searching on that site returns several hits – which are also all 404. Weird! Thus I also found no independent benchmarks.
Crates.io shows a decent amount of downloads. But the version history has it racing through a myriad of micro-updates. Each of those would incur technical (testing…) downloads, quickly driving the total up.
As for the code examples, .await on every tiny step must be overwhelming tokio, no?
If it sounds too good to be true, might it actually not deliver half of what it promises?
It is another HTTP server implementation based on tokio, seemingly developed by a single person. I couldn't find any comparison or description of a different approach, which would suggest it being faster than other, more established HTTP servers based on tokio, like Axum or Actix. Plus the release history does not make me confident that the library is very stable.
You are not the only one that noticed it. It seems to be mostly if not completely AI generated. The dev.to posts also likely got deleted due to similar spam from the same author.