yes

It's perfect! Thanks

It's perfect now !

yep this is nice now, thank you for fixing this little issue, much more confortable!

Perfect

😂😂

Better than my Mother Tongue Fortan

@@SR-zi1pw Not better than my mother tongue Binary

no way :O

Most underrated comment on the internet.

lol

A compiler written in python could still produce code faster than c

@@anderdrache8504 lol xd

@@dextreme1754 he's right though, the compiler has no effect in the speed of the compiled program after its been compiled

ohh, I found a Brazilian here, I'm Brazilian too

Or instead of pushing out assembly, push out LLVM and tap into its optimisers

i3

Я только собирался написать и решил поискать не заметил ли кто другой))

Это debian linux и emacs в качестве текстового редактора

8:00 "моя идея" сказал xD

My uni made us write a compiler instead. Wrote a compiler, mostly using OCaml, from a language called Tiger, down to both LLVM and x86-64. C isn't exactly machine code either. Why stop there?

@@casperes0912 I know this courses, we had to write a Lisp Interpreter in C (i'm more old school, nothing like LLVM existed at my time). But it still does not explain advanced concepts like closures (simple struct with stack variables turned into struct members), script language variable lookup (it's just all dynamic hash tables), pattern matching (syntactic sugar around big nested ifs), coroutines (a new entry point address inside a function + a closure), declarative programming (constraint solving algorith or backtracking). It's terrible how this knowledge is all lacking and people do not understand even simple things and can estimate the cost. I got almost all of my points on stackoverflow with just explaining this in answers.

As a senior software developer having worked with many languages on large-scaled projects and worked with a ton of people both more experienced than me as well as beginners; I hard disagree.

@@dealloc hard disagree on what? that people lack solid understanding of language features or that every developer needs to learn them?

@@dealloc disagree on wht?

This is a forth-ish language. It's purely procedural and stack-based. You read from left to right, there is no real parsing.

Yes, the simulation is also gonna be rewritten in Porth itself.

What's the 'rust equivalent' implementation which takes a full second? The code does almost nothing with memory allocation, so I don't see where the borrow checker would stop you from just copying the C version directly.

can you share the code? i might be able to help optimize it.

@@cookieshade197 i copy pasted the c code in an editor and and translated it to rust (pure, no unsafe block). it took only 0.019s even in my Android (C version took 0.016s). so either he is lying or he just started learning Rust. edit: i removed print! (which does unnecessary formatting) with stdout().write and the rust version is now taking 0.014s. so faster than C :)

@@proloycodes I made a reply earlier but it seems like it dissapeared. I'll try again; the 1 second avg. came from running the optimized binary with hyperfine multiple times. This is an avg. over thousands of runs, but can be skewed since the overall runtime of both binaries is less than 5ms (hyperfine shows a warning for this). I also wanted to post a link to my impl. but I guess that is why the comment was removed.

@@dealloc could you show us your "rust equivalent"? im curious as to why mine is a little faster than c, while yours is taking a whole 1 second longer🤔

yep even i as win user know this

Ctrl-R

The question is whether it's faster to type "!!" or to press the up key followed by the home key. Probably faster as you say

@@Tulanir1 i think !! Offers more flexibility. It is not limited to doing up-up then something else. You could do something before like echo ‘hello’ > !! > grep ‘hello’ So it is, for me, many times superior to doing up up. But I think up up is universal, where it will most likely work everywhere you need it to.

The language is compiled to x86_64 assembly. And in x86_64 we have a syscall instruction www.felixcloutier.com/x86/syscall.html

Font: Iosevka Color scheme: zenburn

Dart is faster than JavaScript and it would've killed JavaScript of Google actually included the dartvm inside chromium but it didn't so the answer is no.

No, soydevs won't be able to understand it.

@@longlostwraith5106 lol

I wish, I spent time learning it so it' better pay off

@@longlostwraith5106 In 2007, Jeff Atwood made the quote that was popularly referred to as Atwood's Law: “Any application that can be written in JavaScript, will eventually be written in JavaScript.”

Better haircut and lighting and he would look good

дааа))

let me guess you are another linux hipster

It depends on your OS, on linux you can do it with mmap. Just don't expect you can simply put some function pointer there and that's going to work. You have to store the raw machine code and then cast it to a function pointer. Just like you would with a JIT compiler, but you also mark the memory as shared

@@mss664 how can I do the storage of the machine code??(I thought that it would be difficult cause the code section of the executable is protected and accecing that part of the memory would result in a seg fault or something like that) I was thinking about c++ delta functions but I am not so sure about how they work, yes linux is my base OS btw

I guess his mother's attic

@@FiveNineO No, it's your mother attic. Gotem!

His talk about assembly may be misleading. You can have the compiler generate assembly code, then run an assembler on that to generate an object file, but compilers usually generate object files directly, which then get linked to make the executable. Likewise with Fortran, or as it was called when I learned it, FORTRAN.

not only loop but the app should require more calculations

@@orkhepaj true

that was the worst, and that keyboard click click

This 'fast enough', 'optimised enough' attitude is why things that used to work great with 64MB of RAM now need 16GB at least.

@@akkudakkupl ... yes, but these 16GB are less expensive and much smaller much more energy efficient than the 64MB back in that old days ... ;-) So that I do prototyping in Python as a MVP (Minimum-Viable-Product) and if the commercial side of things requires it ... I rewrite it than much more optimized in Go ... That way the development process of such MVPs are quick & therefore low-cost ... and if the economic side due to first market responses is promising ... than I invest ... time & money (resources) ... BUT ... I still try to reuse as much as possible components which are stable & well tested ... because all these many available Python libraries are just great in this respect ... and Go libraries are growing fast in numbers & quality ...

flashes sounds cooler

1. He knows the language very well 2. It's fast to iterate with as you're not required to use static typing and deal with memory allocation 3. It's beginner friendly and easier to run and debug 4. ??? 5. Profit

@@dealloc Python is useless just use js

@@dealloc dynamic typing make a language slower than statically typed languages plus they become more prone to type collisions that's why no one use python for development of big softwares (except few)

@@borgir6368 Choosing a language for performance for a hobby project is not what you should spend your time on. Even then, Python does have static typing. See next tsodings next video where he introduces static types into Porth. To your last point, I've got to laugh. You're being ignortant.

no, its just that the C compiler produces more optimized assembly than the Porth compiler

ah, ok thank you I was confused. I misunderstood him probably.

C is still faster that's why whenever you need absolute speed and control to kernel level everyone still use C and that's the reason all the drivers still written using C

Это вопрос или ответ? Если ответ, то он мне не ясен.

@@ggguser потому что больше просмотров?

просто английский - язык айтишников наряду с C)

и правильно делают

А что он это скрывает?