Yes

It's super complicated and i didn't understand most of it. But i liked it bc i wanna learn more about js--> c++ stuff

I loved the video. It's really difficult to give content like this that deep dive into the internals

Duh

Yes please, I really like that the content is aimed towards explanation for intermediary stuff. Especially the fact that you’ve posted a link for further explanation and I can go and see things I don’t understand. There’s already content on KZfaq code space aimed towards beginners and people with an agenda to obscure info to sell a course. It’s why I mostly avoid KZfaq these days and just focus on odin project.

It has done executing while Rust hasn't even compiled!

Or is it infinitely fast?

@@wlockuz4467 I just thought... You have an algorithm written in both JS and Rust (both in source form). You enter commands "node program.js" and "cargo build && ./target/debug/program" respectively and measure time from pressing enter to the program returning. How large the input must be (the N in big O notation) for rust to win? Then check how building for production changes the results.

@@igorordecha I made this test some time ago with python and it really depends on the operation and performance of the machine. Using an 7th gen Intel core i7 and a simples multiplication operation, even with compilation rust executed faster than Python. I assume that JavaScript has a performance similar, if not worse, than python

@@igorordecha cargo run**

What I don't understand is if you're making a special case for infinity why would you run it immediately instead of simply not scheduling it at all?!

@@funnyav "This is the way"

@@ricardoamendoeira3800 As far as I understood, the converting of Infinity to 0 has nothing to do with `setTimeout`. Rather, `Infinity` converts to 0 if you try to treat it as an integer. If you do `Infinity | 0` you get 0.

@@funnyav divide by zero would be NaN. Weird stuff

@@nathanhedglin931 setTimeout with NaN also executes immediately. Fun huh?

thats the point! no need to bag on JS, its been bagged on enough, its more "why" does it do something

👍👍

@@ThePrimeagen Actually I think this was more the "how it happens", in the future it would be nice to include the discussions on why some decisions were made (from issue trackers, PR's, etc. For example why choose to treat infinity as 0 instead of not scheduling the callback at all.

So basically, it is V8 that is shitty.

For me it's not completely the first thing about V8 that I have watched around the subject of data types. I forgot which video it was, but it was some years ago and they had a very detailed explanation about the V8 engine and how you could write your code smarter to become more high performant. They also talked about SMI numbers and how using arrays of SMI's where the most efficient ones to use (because of preventing those extra pointers and memory allocations), also how important it was in general not to mixup several types in the same array. For example 1 object or string inside your array also containing numbers makes the whole thing based on extra pointers and allocations (in other words, the whole array becomes a boxed one). It's also true the language has a ton of initial design problems. But other languages also have their issues. C and C++ have their own huge problems for example (being very unsafe and vulnerable for example). It's great that powerfull JS engines exist these days. I would like to know more about the Firefox JS engine as well. Because V8 is not the only workhorse one that exists.

Facts

@@chrismuga

yayay! its nuts that v8 team is so talented they took a language that is TERRIBLE for perf and made it semi decent. its ackshually nuts.

ya!

​@@ThePrimeagen And make them pretty LongLongLong xD

:) ty bud !!

hmmmm, your last sentence "he is the best Engineer I have ever worked with" sounded kinda sus. Blink twice with the other eye if you need help!

Hey, it was a chance to use the old “you should see the other guy” joke ...

LETS GO!!!

:)

oh bash... stawp

yayaya

I said we as in we made this video together apparently. I don't know why I typed that.

btw, i am not trying to make fun of javascript, just trying to have fun with it :) and try to explain things in a fun way!

classic case of stockholm syndrome

@@ThePrimeagen you killin it, thanks for the good content. 👍

Yo... You love JS? Are you alright? Do you need therapy? /s

yayaya

lets go!

LETS GO, KEEP COMING BABE!

hio! ty bud :)

hey thanks stefan :)

oh well ty :)

i think you have achieved a victory?

i feel special sometimes

ty

I had a look at the MDN page (probably the most readable description I could find), and apart from warning about non-numbers being implicitly coerced into numbers, it says nothing like this. There are situations where delays may last longer than expected, but they should never be shorter. So it seems like this is some kind of weird Chrome/Chromium bug.

@@lawrencedoliveiro9104 MDN is not a specification… I believe ECMA-262 is the authority for this behavior.

I found the answer: the W3C HTML5 standard says “if timeout is an Infinity value, a Not-a-Number (NaN) value, or negative, let timeout be zero.” So this is behavior is by design.

@@Vim_Tim But that doesn’t say that large values should overflow.

It it interesting to consider edge cases like in this video, but hopefully this issue shouldn't arise too often.

hopefully in a good way

@@ThePrimeagen HAHA right

its crazy out there :)