Hacker News new | past | comments | ask | show | jobs | submit login

Hermes itself is already faster than JSC, which is on par with V8. This just puts it way ahead and probably closer to Swift/Kotlin code, at least in the realm of mobile apps.



What data do you have to back up that statement?

We've seen the opposite for networking related activities especially when there are 100's of requests that can happen on certain screens. We've tried using Hermes but didn't see any differences other than the app bundle increasing in size.


No, the Hermes interpreter is probably on-par with the JSC-Interpreter if this is a comparison made on iOS React Native apps. JSC-JIT is usually on par or ahead of V8, the iOS JIT restrictions really fudges up results since it's so imperative for JS performance.

From what I've seen Kotlin(and Swift prob also) semantics are fare more AOT-compile friendly so they're probably a fair bit ahead (depending on how much you gimp your JS target to mimic Asm.JS/Wasm but most regular JS developers won't write code that way).


The Hermes interpreter is probably faster than the others because (afaik) it interprets bytecode produced by the hermes frontend parser (which is don’t ahead of time).




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: