Message from JavaScript discussions

October 2018

— Now the results are at the mercy of your OS scheduler :P



Make systemd start the tests immediately after the OS loads on a very minimal setup, and restart the device twice with different cases

— Or build and run a single purpose OS that boots, runs test, prints, and shuts down

— NodeOS on bare metal

— Oi oi🤤 chat +100 messages, well done. is Thomas dancing on my bones there. i will rise later after reading all that..

Message permanent page

— Averaging would be easier? :P

— You can solve this by running the process in a realtime OS I think

— That would lead to the exact same amount of time every time

— Restart hundreds of times, calculate average by hand

— Why by hand?

— Arrow functions may be EXACTLY the same, so they (developers) reused the part of normal implementation to arrow, okay. They can't be MORE performant, it's imagination and some measurement errors. SpiderMonkey in FF v40 had 70% degrade on arrow functions, i googled some info about, so they fixed it. Good transpiler will always be ahead of that. Is arrow function super-duper different that bound function? No. So, not a big deal. Just a syntax.

Message permanent page

— They ARE more performant in SpiderMonkey at this very moment...