I switched my blog engine from Warp to Axum just a few days ago, and what you said in this video pretty much echoes my feelings. Axum is just nicer to use than Warp imo, and things like sharing data are much cleaner. And I’d definitely recommend the debug_handler macro, it’s seriously helpful when you have issues like holding a reference across an await call, something that’s otherwise very hard to debug.
hey! im super new to rust but the one I was thinking of using once i get comfortable enough to was rocket because it looked so easy. Is axum a better choice than something like rocket?
You should’ve just used Tokyo for for receiving the asynchronous requests, and then deserialize those requests with serde. You’re gonna get much faster speed in a much more stable and transparent API imo. Now I’ve never done anything on the scale you’re doing it, so there may be a reason not to, but I really can’t think of one. Serde parsing should scale linearly. If you don’t know what I’m I’m talking about, you probably need to dig deeper into serde.
So happy to see the talk on Axum !! Thanks a ton guys for sharing the content !!
I switched my blog engine from Warp to Axum just a few days ago, and what you said in this video pretty much echoes my feelings. Axum is just nicer to use than Warp imo, and things like sharing data are much cleaner.
And I’d definitely recommend the debug_handler macro, it’s seriously helpful when you have issues like holding a reference across an await call, something that’s otherwise very hard to debug.
hey! im super new to rust but the one I was thinking of using once i get comfortable enough to was rocket because it looked so easy. Is axum a better choice than something like rocket?
I was going to ask same question, Rocet or Axum 🤔
Rocket is a dead framework, abandoned
You should’ve just used Tokyo for for receiving the asynchronous requests, and then deserialize those requests with serde. You’re gonna get much faster speed in a much more stable and transparent API imo. Now I’ve never done anything on the scale you’re doing it, so there may be a reason not to, but I really can’t think of one. Serde parsing should scale linearly. If you don’t know what I’m I’m talking about, you probably need to dig deeper into serde.