Despite that, it's nice to read a positive feedback :)
Despite that, it's nice to read a positive feedback :)
In either case, it's not using node or v8/chromium .
Node.js adds various standard native modules - mostly various IO (file, tcp etc...), a way to load other modules, access to the underlying OS (environment variables etc...).
running v8 alone is akin to running the browser without html or the dom api, you basically can compile and run javascript functions, but can't do much with it.
The Julia websites has a benchmark for various languages performing some tasks. Javascript seemed pretty comparable to Java.
1. c/c++ is the fastest.
2. modern java is either the same as c/c++ or about 2x slower
3. modern javascript (i.e. v8) is about 4x slower then java.
4. python/ruby and most other dynamic languages are about 10x slower then java.
Never seen C# really compared with others because it's rarely ran on the same systems.
Go, Rust and all the new shiny languages are still far away from when we can infer about their performance, but they are probably somewhere around the Java level.
Didn't mean to imply that other shiny languages are in the Java level, just bad grammar on my part I suppose.
So just a hunch, also some barely usable benchmarks that appeared on HW a few months back.
What gave you this impression? MIR landing?
I'm not saying it is a wrong impression, but it doesn't match mine. My impression is that Rust is routinely very fast, often rivals C, and most of the recent headways were surrounding the performance of the compiler, not the resulting programs.
You seem to have cherry-picked that single-core-x86 n-body measurement just because the quad-core-x64 n-body program measurements are slower than the Java program measurements.
http://benchmarksgame.alioth.debian.org/u64q/performance.php...