Is this language a good or a bad meme?

Is this language a good or a bad meme?

Attached: IMG_20180405_170053559.jpg (2500x3333, 620K)

Other urls found in this thread:

insights.stackoverflow.com/trends?tags=rust,go
twitter.com/AnonBabble

is it better than c

Attached: dafoe.png (1029x994, 1.08M)

Meme. C++ is the only God-tier language

But it is trying to do things better that are kind-of shit in c++. I know that it will still take quite some time, but the outlook is quite good, r-r-rite?

No.
Truth.

friendly reminder that c++ if fucking garbage

Not only that, it doesn't have a garbage collector to clean up after its excrement.

It seems to work very well in large projects judging by the pace of progress in Servo and Redox OS. It's not mature though, and many language features still need a nightly build of the toolchain. Once that is no longer the case I'll take a serious look at it for my own programming.

>It seems to work very well in large projects
But it's never been used in a large project.

I wonder if the borrow checker could just be an external linter working on AST or some other IR to simplify the toolchain a bit.
It doesn't target what C is good for - highly standardized language with good support across systems, architectures and compilers with good bindings to other languages.
If it can't beat Fortran with performance then that would be pretty disappointing since the mutability and ownership scoping and type system is a massive hint and guarantees for compiler.

Tea

It's still got a long way to go, but it is the only serious possible contender (not even a replacement, just a contender) to C in the area that C is good at. It does seem to solve many problems that C has, and has quite good tooling, everything has been thought of.

There are problems with it, and I am not an expert in Rust (I am with C), but it is certainly a language with a lot of potential.

The main problems with its adoption in the embedded world is the fact that it is quite a conservative one when it comes to languages, and even more pressing is the fact a toolchain may not be available for the platform you want to target.

Time will tell, but the language has potential, and as I have said - it is a *contender* in the same arena as C, unlike Go, C# or Java (which are ok languages for their niche, but not meant for systems or embedded work).

Oh boy.

mozilla
>"hey guys our browser has memory leaks"
>"fuck C++ let's write our own hip and trendy language so we can rewrite the browser"
>5 years later
>compiler and standard library still buggy as fuck and suffering from feature creep just like C++
>servo still not even remotely usable

google
>"hey guys our browser has memory leaks"
>"let's write a bunch of compiler tooling"
>contribute everything back to both clang and gcc
>every single C/C++ program can benefit
>bugs squashed everywhere
>browser immediately sees massive improvements
>even bugs in firefox are uncovered

the sad part is mozilla was probably forced to do this at risk of being further cucked by google

Golang is the only serious new systems language because you can actually write programs with it.

>rust
It's been what, eight years now? Golang has been around for about as long and has already produced tons of programs that are damned near essential for modern software development. Couldn't live without docker for quick and easy test setup.

>b-but muh fizzbuzz doesn't need such things
Neck yourself.

Writing Rust feels like rolling in shards of glass.

But why is it the most loved language on stackoverflow? Serious quesion. Not even memeing

because code monkeys are too stupid to learn haskell

Why is an unfinished, unready language like Rust that promises memory safety and to "do everything better" "loved" you say? Because just like others have said, Rust is a meme.

Rust is not the most wanted, nor is it the most used (I assume we're thinking of the same survey). That says it all.

one is a systems programming language and the other is more or less for math. So pretty different fields.
>Rust is not the most wanted, nor is it the most used (I assume we're thinking of the same survey). That says it all.
Yeah sure it is not the most wanted, because it is not standardized yet. That might also the reason why it is not the most used. But it has a pretty good outlook, right?

Your moms ass also had a great outlook. Look at it now.

0/10 Pajeet. Shit in street.

Because programmers like new toys, and Rust is a cool new toy. No one wants to learn C/C++ because it's old and boring, while Rust is fresh, new and open to possibilities (it will never actually accomplish)

>(it will never actually accomplish)
But why do you think that?

any language other than C, Assembly, Ada and Python (for doing a quick project) are unnecessary. literally every single thing that exists today could be made with these 4 while making it(a program) alot better. creators made all those other languages just because they could, not because they(or anyone) needed to.

Because it's a new thing and only old things are good

languages*

There are only two (three) languages you should be learning right now: C/C++ and Javascript. All other languages will become obsolete in the next 10 years.

>Java
>Obsolete
Pick one.

>> An OS is not a large project.

Attached: Redox_running_Orbital.png (1920x1080, 1.12M)

Lol Java is but obselete. Oracle does nothing but make good things obselete and they sue companies for using their crap.
Also,
>Java can't even cross-platform console beep

insights.stackoverflow.com/trends?tags=rust,go

Just a toy os, literally millions of those around.

idiots have taken over the whole c++ steering committee and have been pushing features that are useless to the language and seriously, almost everyone I know who writes c++ doesn't use the majority of the features introduced after c++11.
I use c++ in my multimedia projects... well in anything that doesn't need to be C, and I fucking stumble upon horrible nu-fag code all the time.
There's the people who write code to brag about their coding skills in new features, but have never been in a 50k++ code project,
and there are people who have to pass their code to junior developers so they stick to the plain OOC c++ without the gimmicks.

BUT CAN c DO THIS?
fn main() {
let v: Vec = (1..=10).collect();
let s: i32 = v.iter().map(|x| x.pow(2)).sum();
println!("{}", s);
}

Attached: 1503739800080.gif (630x354, 3.58M)

>443MB
S-so this is the power of Rust.
I will look at the code and rewrite it in C for efficiency.

Attached: stages_of_denial_second_edition.png (1024x1024, 338K)

Hello 2028.

Please include me in screencap.

it's like more autistic ruby

>google
alphabet shill detected.
google buys android, android is in java, android remains in java.
google forks khtml/webkit, it's in c++, remains in c++ .
Let's see what a google project which started in google, looks like,
>Fuchsia
>written in C, C++, Dart, Go, LLVM, Python, Rust, Shell, TypeScript
Dart = developed by google
Go = developed by google
Python... pick a slow interpreted pos because we didn't have more than 5 of those
TypeScript... we didn't have time to implement our own JS so pick any pos that doesn't say JS in the title

Rust... well, spank my ass and call me nigger
>> compiler and standard library still buggy as fuck and suffering from feature creep just like C++
where's my lambda metaprogramming templates with functional fries and ketchup.
REEEEE I cannot program without features.

#include
#include

int main(void)
{
int result;
for (int i = 0; i

>I wonder if the borrow checker could just be an external linter working on AST or some other IR to simplify the toolchain a bit.
It might be possible to do this with C's pointers simply because of how few types C has, and I wouldn't be surprised if some researcher has a linearly-typed dialect of C. Having said that, it's really more useful in a language like C++ or Rust where a bunch of standard library types use RAII behind the scenes, but I don't think it's realistic for adding to C++. It would be tons of effort to encode every standard library type's behavior. You'd have to in effect recreate Rust's trait system to manage behavior that isn't just linear, and then you'd still have to add traits to every third-party library. By that point, you'd have reinvented Rust more poorly.
>It doesn't target what C is good for - highly standardized language with good support across systems, architectures and compilers with good bindings to other languages.
This is a little less true now, but sadly you're right.
>If it can't beat Fortran with performance then that would be pretty disappointing since the mutability and ownership scoping and type system is a massive hint and guarantees for compiler.
It could BTFO out typical C++ with aggressive intraprocedural analysis given a hypothetical more mature compiler because of all the type info, but Fortran is entirely a language of procedures and lacks pointer indirection/boxes, so its primitives are kind of inherently faster. On the other hand, people don't write operating systems in Fortran, for good reason.

>Golang is the only serious new systems language because you can actually write programs with it.
Golang is garbage collected. You can't forge an address into a pointer. Calling Golang a systems language is kind of like calling Java a systems language. Like, if you squint really hard, maybe. Rust is actually competing in a space similar to C++.

IMO, after working in embedded software for a while you'd end up needing to use unsafe so much for half your programs you'd wonder why you didn't just use C anyways.

Basically what I see is in the embedded world the problems rust aims to solve don't really affect your project enough to justify rewriting all the tested and working code in rust.

>IMO, after working in embedded software for a while you'd end up needing to use unsafe so much for half your programs you'd wonder why you didn't just use C anyways.
I don't write embedded but I do feel this feel user. Having to write Unsafe Rust everywhere confers no benefit over C (apart from the Rust package and build tools). I hope quality-of-life enhancements for people not using the standard library (maybe even a more spartan standard library for people doing embedded) emerges.
>Basically what I see is in the embedded world the problems rust aims to solve don't really affect your project enough to justify rewriting all the tested and working code in rust.
I don't think rewriting fully functioning C codebases in Rust is a good idea period. The only people I've heard that unironically advocate this are idiots.

>no benefit over C (apart from the Rust package and build tools)
Those build tools aren't a benefit. They're a cost.

I really like the language, and I think that my work would be made much easier if better support for embedded targets were available.
I think I'll wait a couple of years before trying it again, though.

I thought that crab is supposed to be a sideways Xenomorph alien for a second.

It's not enough for a language to be better to secure market share and feature_stuffing != better. C's elegance comes from its simplicity. It's here to stay for the foreseeable future.

Today for the majority of work above the C level i would use Golang. Simple to learn, fast, GC'd, better memory usage than Java (and speed), great standard library, concurrency out of the box, cross-platform, super fast compilation time.

C > Rust

oh don't get me wrong, google is shit and most of their projects are garbage, but this is the one case where their compiler guys did the right thing instead of just memeing hard like mozilla

>where's my lambda metaprogramming templates with functional fries and ketchup

it's in lisp which nobody writes because as it turns out, lambda calculus by itself is not sufficiently high level enough for the average corporate drone programmer

*C

Golang binaries are an ugly garbage piece of bloat.

Rust is simply bad. Don't bother.

It allows for a high level of abstraction, with safe concurrency, and performance at the level of C++, and that's only with a fraction of the compiler performance work of C++.

Basically the bonus of it is that it makes developing with open source a lot easier, and it makes doing concurrency easier, meaning programs written in it will actually utilize all of the cores you have, and because it's designed to be memory safe, means less debugging, easier maintenance, and a lower barrier to entry for various projects.

Basically, it's the best shit ever for FOSS.

If you need to use unsafe all the time, then you probably don't know how to use Rust effectively.

New solutions and paradigms mean you need to rework the way you attack the same problems. If someone gives you a chainsaw, you don't try to use it like an axe.

Garbage

lol no generics

When your writing up kernel code to handling interrupts and other asynchronously driven code/drives rust's advanced saftey features offer no improvement over C. You have to use the unsafe. If you look at the redox github you'll see this all over the place.

At least as a person who does embedded kernel development I see no reason why to use rust. Maybe for c++ apps, but I don't have much experience with that so I do not have strong opinions on the subject.

Attached: Ev8ni.jpg (530x706, 142K)

>I've run out of actual arguments.

Rust really really really really needs to ease up with its borrow checking. Maybe even provide reference counted smart pointer for opt in automatic memory management and allow multiple mutable references.
Other than that Rust is pretty damn solid.

Multiple mutable refrences with Rust would require them to rewrite the entire stack orientation they have blessed garbageless.

>Rust really really really really needs to ease up with its borrow checking.
You'd have to be specific, but getting rid of type safety defeats the purpose.
>Maybe even provide reference counted smart pointer for opt in automatic memory management and allow multiple mutable references.
This already exists in the standard library for precisely the use case you have in mind. Look at Rc and Arc.