Nim in 100 Seconds 

And also in the less ethical hacking space ... Some virus scanners even blocked many Nim programs just because they looked like they've been written in Nim.

@@dariusduesentrieb Don't blame hackers, blame the antivirus mob using it to squeeze money from you. Many of these are excuses to force developers into paying for whitelisting their products...

Cool, never even heard of it before.

and vlang too

I wanna learn ethical hacking. Can someone plz give recommendations on how to get started? Thanks in advance

Love it when small companies are not afraid of using something unique. What do you guys do?

Was just gonna ask what’s the startup called?

Some of that tooling you're referring to might be another market. Wouldn't mind seeing what that's about.

Embauche moi j'en ai marre de typescript tout notre backend tourne sur node c'est un toaster géant 😭

@@alexandredaubricourt5741 ça devient toujours le même quand les projets avancent

Looks awesome but the complexity can be overwhelming, is it a good Solo-Only programing language or do you have a good experience with tenths of programmers working in the same code base?

I wish Nim had the backing and man power of Rust, then it could be something great.

When I'm investigating which programming languages big companies are using, I always found some unpopular languages on the code base. There's a reason why seniors decided to use them.

It's completely useless

​@@nicocesar I use Nim as the main programming language in my company here in Paris. Team size is 6 devs, junior and mid-level. Our productivity is now insane. I have to confess I had to write some tooling and utilities for them to use Nim easily. That said, we use Nim now for everything, from Really Big Data to Web Development & ORM, low-level RPC stuffs, to fine-grained AI/GPU computing (thanks to Arraymancer)

ah ok. clumsy to have chose let...

And types declarations of Pascal / Oberon / Modula

There's also Julia

@@happyjohn1656 Julia is terrible for most purposes. Sure, may have c-like speeds, but it also has c-like compile times that slow you down every single time you run your program

@@joeymea julia gets nowhere close to C except for their cherry picked array benchmarks

@@joeymea Fixed that for you: Not C-Like compile times but C++-Like compile times.

This has the be the best "first" comment I've ever seen. bravo

Okk you got the -first- 0-th price 🏆

Congrats

Phew! This made me think that this programming language is not 0-indexed!

RU-vid commenters are evolving lmao

The nice thing is that the python interop, if needed, is pretty easy. Want to do something computationally expensive but there isn't already a lib out there that wraps c-code for you? Simply write it in nim, compile it and with pretty low amounts of effort you can use it in python through a package called nimpy.

@@isofruitfruit9357 oh wow that’s awesome!!

I'd want the opposite, using popular python packages in Nim. Getting that part of the ecosystem would be extremely valuable

you can read the docs

@@crackwitz that would be fantastic as well

That's right and Nim uses both with precisely this distinction and it's enforced by the compiler.

Beautiful

I don’t understand this well but I want to. What’s the difference between an impure function vs a pure function in the context of async/sync?

@@imelbertkim pure functions are, by definition, functions that have no side effects and whose output depends only on their input. This has many advantages, as you are guaranteed to get the same output given the same input every single time. Therefore, parallelizing the calls of pure functions is trivial, as you don't have to worry about deadlocks, race conditions or similar issues.

@@henriquefinger935 thanks for your explanation!

people like python's indentation?

@@maxwellwellman Yes?

@@maxwellwellman Yes.

I'm not a fan of their styling conventions though.

And macros from Lisp

Wanna try a couple of new ones? Gleam, Pony, F#

@@matthiasschuster9505 I use F# pretty regularly at work. I've looked at Pony but it's essentially dead. Last I checked all the maintainers are working on Verona or something like that. Gleam runs on the BEAM, right? I've done a small phoenix app in elixir and hated it... which sucks because I wanted to like elixir. It may have just been phoenix that I didn't like, though. I love the idea of actors and that approach to concurrency and would like to build out the idea more in F#. Agents are cool but nowhere near as robust as what Erlang/Elixir, and I assume, Gleam have.

It's about to !

Go and Rust has billion $ companies pushing them. All NIM has is a community and the fact that it's an amazing language. But NIM is literally SO good that it will become mainstream even against these odds. Personally, I switched and now only use NIM in all situations where I'm not locked to a language because of a platform. Even my beloved Python 3 has been retired 🏖 Projects already written in Python, I slowly rewrite in NIM as I continue development.

@@ViaConDias What about the lib ecosystem? Is it good? Like dumping python you also dump all of that right?

Nim will take off because I am looking into it and the masses always follow my thoughtform!

@@danvilela there're plenty of Nim libraries, many of them native. And many are wrappers of existing C libraries, if you can't find a library that suits your needs Nim has excellent FFI with C, and it's even possible to call Python libraries with `nimpy`.

I_did_not_see_underscores

Fk... this guy reads our mind (or probability to blame?)

@@vaisakhkm783 Frankly it's happened too many times to be probability at this point. There was one point where he put out 4 videos in a row that I had been looking into the days prior...

Nah, Jeff is clearly psionic with telepathic and precognitive abilities

@@vaisakhkm783 Nim was definitly on his TO-DO list. At least bc he mentioned it in his Iceberg Vid.

You can do so much insane stuff during compile-time. In webdev I used it literally to evaluate in an ORM whether the Models you gave me for a many-to-one query were correct (so if the "Many" side had a Foreign-key field to the "One" side of the relationship), allowing me to give users compile-time guarantees. If their code compiles, their query for the many-to-one and many-to-many relationships will be correct and work, no runtime errors in that regard.

@@isofruitfruit9357 Fascinating; compile-time guarantees on correctness sounds like a really powerful capability - the only other language I heard could do something like that was Haskell.

In Rust, the sqlx crate does something similar with macros - you write SQL queries as strings, sqlx calls the database at compile time and checks the returned type. Haven't used it much though ^^

@@saeedbaig4249 You can basically force anything to be used at compile time with the {.compileTime.} pragma. I often enough do this to do some arbitrary logic checks here and there, SQL is the most prominent example I can come up with since in an ORM you're rebuilding a schema with inherent logic connections that aren't going to be otherwise represented in the code. Nim has another feature called `Concepts` which are a bit complicated to use imo but are essentially interfaces that allow you to check types at compile-time for certain properties or behaviour, such as whether a method with a specific signature is defined, whether a method returns a specific value for them or similar.

Good idea but bad design and bad implementation. Good for throw away prototypes though.

@@xnoreq Hard disagree here. The implementation perfectly useable in general. Sure could be better, but claiming it's unuseable for production is not correct.

@@isofruitfruit9357 Disagree all you like, it's still correct. When a simple hello world turns into 300 kB of JS and you turn to the main developer of the language and he says "it doesn't matter" plus all the other issues the language/transpiler has, good luck using it in production. I'd fire any of my devs for grave negligence for using it in production.

@@xnoreq Again I find you here, spreading lies, I have a web app with karax Nim's frontend I am developing with 2 out of 10 pages already implemented and its 500kB. Yes when it's done it would be a lot bigger. Now I just compiled a hello world and its 137B!! here is the code: proc kout*[T](x: T) {.importc: "console.log", varargs.} kout cstring"Hello World!" Get a life!

Generated code: /* Generated by the Nim Compiler v1.9.1 */ var framePtr = null; var excHandler = 0; var lastJSError = null; console.log("Hello World!");

Amen

The power NIM gives you is second to none so you'll learn to live with the syntax

it is pretty beginner friendly and optimized to write software pretty fast, i do not know why that has not been a deal before, the verbose languages like java are a hell to me to write something that does something without spend my live on it, i like python and lua syntax because i can write a lot and do a lot of things without have to worry if i have a syntax error, i can focus on logical errors more than if a symbol got out my way at writting. Well at least that is my opinion after programing in java and in C/C++, python and lua.

Java is indeed hell

hello world...

@@LeonLibertario A chess engine

@@dariusduesentrieb oh really okay bro nice

He makes 'codereports'

Making 100 sec video that requires 5 mins of research on wikipedia is not "keeping up to date" with anything

Note it can also use existing python code trivially via nimpy and vice versa can trivially used to build libraries for python via the same nimpy

It has! Better ask question in the #embedded channel in discord.

On that note you may want to look into the nim-based Ratel package for microcontrollers that can produce some absurdly small binaries. The author of said package is active on said discord from time to time.

@@antonisgeralis5266 oh cool! Is that a Fireship discord? Or just a generic embedded discord?

@@evanbarnes9984 nim's discord (sorry can't post links in youtube)

+=1

Uncautch error: #####

However, tbf there is just a lot of features that he could use here and he was already running over 100 seconds. Stuff like easy use of templates, easy compile-time evaluations, the ability to have style-insensitivity (so if a package in your dependencies uses camelCase you're not forced through their package to have camelCase in your codebase), how you can modify all that behaviour with the compiler, concepts and more Just a lot to cover and not a lot of time

Similarly for you, the appeal could lie in nimpy. Sometimes you might run into scenarios where the implementation in python has a particularly expensive step where there is no wrapped c-lib for. You could write that part in nim, compile it with nimpy as a python library and then use that code from python without issue for a nice speed bump.

@@isofruitfruit9357 That sounds really nice. I will definitely look into that. Thanks

You can create stand alone executables with Nim. Tools that friends need, just send them one compiled file. It is different from python entire environment.

In general compilation is pretty fast. Think faster than C++/Rust and without any multithreading, or other optimization. The nim compiler compiles pretty fast (12s) in a decent machine and uses about 600mb. However huge projects suffer from slow compile times and there's work done towards incremental compilation in order to fix these issues.

I do not use C so I have no basis for comparison, but basically for debug builds with around 11k loc on an intel 11th gen i7 I take less than 10 seconds (and that is with the compiler copying a bunch of generic code around a lot), so I'd say in general the compiler is fast enough.

@@antonisgeralis5266 wait, nim transpiles to c++ and then to machine code, and it does all of that faster than c++ to machine code??

this one specifically was begged for over the last 12 or so videos posted in this channel

but if the official page says that it goes for version 1.6.8

@@dgames218 yes but the default memory manager is confirmed to change to ORC for the 2.0 version. The same mm can already be used in current (and a few older) Nim versions

Obligatory note though to keep in mind that any kind of synthetic benchmark in general is pretty much meaningless. Like, the most prominent HTTP server in the nim space, httpbeast, does perform well in the http-server benchmarks I've seen (as do servers behind Artix, Rocket and boy will the performance of some of the JS-based servers surprise you), but the benchmarks themselves are fairly far off from real life scenarios. And It's been often enough discovered that benchmarks keep comparing wildly different implementations across languages, such as the implementation in one language being multi-threaded while the other is single-threaded for a highly parallelizable workload. The key thing to keep in mind is that you have low-level memory control and *can* tweak stuff as needed to get to the performance you need. How much time you end up investing for that depends on the individual usecase, but generally I've stopped looking at performance comparisons between languages that give low-level control. Once you have that it only depends on your knowledge on how to optimize your memory layouts, copy behaviour and the like.

Plainly false, you can see Julia at the top of the benchmarks games, along with C, C++ and rust.

@@12nites those benchmarks and Julia’s own micro benchmarks are the only ones to show that.

I never spend any real time with Julia so I can't comment, but the thing about NIM is, that it's so simple and beautiful that I never need a higher-level language, and it's so powerful I never need a lower-level language.

Nim has multiple dispatch too

@@offensivearch Thanks, I started using Nim about 2 months ago and I loved it but I also missed multiple dispatch and couldn't find info about it in the docs. Now thanks to this comment I looked deeper into it and learnt that in Nim it is called dynamic dispatch, and the diference between `proc` and `method`.

Personally I use both Julia and Nim depending on the project and I love them so much!

Nope. Just tooling, debug story, and lack of libraries. Takes a while to understand errors. The language itself is perfect, of course from a totally objective view ;-)

@@kobi2187 Because building wrappers around C headers is so easy, don't you technicaly have all off the C libraries

Nim was actually called Nimrod until 2014.

you mean compile time?

Runtime? really?

@@TheMR-777 my bad compile time it is

@@tjgdddfcn compile time