IT'S HAPPENING !
Microsoft BTFO all current programming languages and endorses rust !
>Microsoft Security Response Center endorses the use of Rust for safe systems programming. :
msrc-blog.microsoft.com
for many years, looks like languages fail miserably @memory safety, it has been constant at 70% since forever.
looks like this is the language for the next 40 years indeed :
youtube.com
>Is memory safe
>is also a cumbersome and unproductive language
Ill pass
This is good news for Microsoft employees as they will be set back to late 90s levels of productivity waiting hours for long compiles, and will be able to openly browse facebook and play as many video games as they want during office hours while they wait.
WTF I LOVE COCKS NOW
better than million dollar segfaults !
also
>Incremental Compilation
doc.rust-lang.org
Fixing all those memory leaks sure is productive!
Well i dont know what to use this language for
Rust does have going for it that out of all the newer "memory safe" languages that have come around lately, its really the only one suited for actually doing low level work like writing a kernel with.
Is there a madman out there that's going to write a unix-like kernel in Rust? I fucking hope so.
Imagine Windows as it is now, but it takes 100x the time to compile and is 10x as large. Not that I care, I don't use an operating system made for children.
in the works
redox-os.org
Oh my god I can already hear Jow Forums screaming in agony when this is released holy shit.
>be a cumbersome and unproductive programmer
>blame the language
It's an experimental OS. They've already stated it will never compete. And the code is a fucking joke.
Haskell + Rust masterrace?
Based Microsoft, savior of the programming world!
>And the code is a fucking joke.
Point out some real issues. The code is widely considered to be excellent and several of the primary authors were hired directly onto the Google Brain team which is incredibly rare considering their age. I bet you can't name anything substantial since you're probably some bikeshedding clown.
No, referential transparency is a meme and focusses on the wrong problem.
Just use nim.
>GC
fuck off faggot
Nim and Crystal are irrelevant now and will be even more irrelevant in a few years. Doing nothing would be a better use of my time than learning Nim. The days where some halfwit can create a widely used programming language are over and Python will probably be the last popular example.
did you read the fucking article ?
Microsoft just said if your language doesn't have a borrow checker you're either retarded or a masochist !
Another strange toy with an uncertain future for the Pajeet-H1B-Female era.
C/C++ will never die.
Anything you can use C or C++ for... It's a turing complete language that compiles to machine code.
Yeah, how's the interoperability between both? None? Oh.
So why are you using linux with all it's shiny shiny customizations? Or do you need a bare bones OS that just boots for your pentium II?
Arguments against Rust: Unsafe blocks with /* dude trust me */ comments are rampant even in the standard library. Memory safe code is implemented in many languages, and they include relatively old and new languages from Ada to D that are fully capable of performant low level systems programming.
Rust has traction and hype. Ada is confined to its niche and the image of being uncool and old, D is confined to a very small and certainly not growing set of fanboys.
C++ is the same, except unsafe is default.
If Microshafts cared about safety it would have used Ada and SPARK years ago. There's more to this announcement than meets the eye.
>There's more to this announcement than meets the eye.
It means that the rainbow sock brigade has infiltrated microsoft too.
i don't think redox is sufficiently unix-like to run much of anything
>it's another freshman-pretends-to-use-formal-methods-because-logic-teacher-mentioned-it-once-in-class episode
Thanks god I'm a CE and not forced to suck trannies dicks and this god-awful language.
Sorry for you CS guys, seriously
>tfw my job is about ANSI C and FORTRAN
thanks god.
> single blog post from a web developer in a company employed tens of thousands of engineers
> same thing as decree from CEO
uh huh.
Please rewrite Windows in Rust. Please god please.
Also, someone rewrite Linux in rust.
imagine the smell
Jokes on you none of those languages are taught anymore.
...
have sex
enlarge the hole
It's a toy project, slow as fuck, it has about the same chance of uptake as Haiku OS (which is fast as a rocket by comparison).
>reading comprehension
why are rust shills so fucking retarded, anyways?
you would think that people trying go get people to use a new lang would actually make sensible arguments.
They just rant about how the borrow checker enforces memory “safety,” as if needing to fight your compiler is not only a positive, but a make-or-break language feature.
All while ignoring the fact that rust is unsafe as fuck and unstable, and will not get formal verification tools - meaning its promises of safety are worthless - that actually safe languages have for a long-ass time, if ever.
And the whole narrative about C++ being responsible for some dearth of vulnerabilities is just absolute nonsense, C++, especially modern C++ is not the cause of memory safety issues.
That’d be C.
C++ actually has a fairly “normal” amount of CVEs iirc.
It took decades for C and C++ to get good sanitizers and they still can't provide all of the guarantees Rust can out of the box. Formal verification of C is hard, C++ is impossible. Neither are "safe languages" either unless you only work with a strict subset like Misra.
Even actual "safe" languages like Ada Spark require language extensions beyond what any C/C++ developer has probably dealt with to be fully verifiable. Have fun writing a dozen lines of pre and post conditions on every function definition.
The F-36 is written in C/C++. Can you say the same about Rust? Lmao.
Um, it's just a blog post. Doesn't mean anything. Most likely this just means more blogs about Rust (still good news for Rust).
Most Ada code in industry is still Ada 85 which is before it got pre/post conditions, and spark is relatively new language considering it's age.
Look up "dearth".
Incremental compilation is not that fast. Cranelift isn't either. I'm skeptical that rustc will ever actually be fast, unfortunately.
What the fuck, I swear it meant the opposite.
Who do I blame for this?
>Rust confirmed pajeet language
Nice. Even more reason to avoid it.
There's also the simple fact that memory bugs are at best a plurality of bugs, even critical security bugs or stability bugs. Rust is not a silver bullet. Is constant programmer overhead a better savings than marginal theoretical differences in QA overhead and the marginal theoretical difference in risk of bugs exploding into security breaches or costly crashes? It's not clear to say the least.
nim is going to have borrow check.
Use Lisp.
No, little Jimmy. Your favorite academia meme language isn't going to dominate industry in 2019.
Using C is a constant developer overhead. Hunting for bugs are runtime with C which would have been caught at compile time with Rust is overhead. Rust also protects against many concurrency bugs which C has zero support for preventing.
modularity fixes most of the compile time problems
redox-os
>a bunch of Pajeets are recommending Pusst
So...don't use it. Got it.
Here's a serious question for Rust shills. Let's assume for a moment that the borrow checker is a good idea (because it is, despite what the brainlets with an agenda on this board believe).
What is preventing other, more popular languages from just stealing the feature? D's already working on a design for one, and I imagine other languages will follow once they realize that it's actually not terribly difficult. When they do, where does that leave Rust?
I don't get memory leaks in the first place.
>no lambda
Not needed.
>void pointers
Literally nothing wrong with this.
>no generics
Who cares.
>MUH OOP
Objectively a good thing.
Seems like Rust is a massive crutch to shitty programmers, if anything.
im not a rust shill
but this is a situation where "the first wins everything" may be true
rust has a cachy name, not a worn out reputation, and hype
Neither does anyone using Rust.
if something better comes along use it, it's hard to bolt stuff like this optionally to an existing language though
Yeah but unlike the Rust people, I don't spend a significant amount of time battling the borrow checker.
I don't "battle the borrow checker". Sounds like you need to get better at programming.
omg, that's the best comment of this thread in which I'll retweet ;D
You can say git gud all you want, but the good programmers will always be outnumbered by the shitty ones, and even the best ones make mistakes sometimes.
They already use C and C++, They are used to it.
Its hard to retrofit into an established language because it impacts everything. It could also be very costly in more dynamic languages.
Much easier to solve compile time errors than hunting down runtime bugs in valgrind.
Rust's advantage is not that they were first, it's that the project is backed by a large non-profit and are using it in a popular widely-deployed application. The project has a lot of free advertising and shilling because of the novelty of the borrow checker, but the project is also incredibly well run compared to its peers. Too many other languages don't have a clear roadmap or keep getting stuck knee-deep in technical debt of their own making, or they have a hostile and insular community that is busy bickering with each other when they're not leering down at you from their ivory towers. Or they just die because they never get that critical mass of interest to get into the mainstream programmer consciousness.
Thing is, Rust is trying to be the next C++, and I think that it's actually decent odds that it could gain a significant foothold in spaces where C++ excels. What I want to know is what will be the next C. Rust is too complicated for that niche, and there's no obvious company or non-profit backed language on the horizon.
The fact that if you implement it, it will pretty much cause a lot of code to fail to compile. So then it will just be an optional feature you can turn on which no one will. C++ already has a lot of the safety features of Rust but it is all optional and requires the programmers to know to use it.
>next C
I don't think there will be. The advantage of C is being ubiquitous and simple enough to implement for the shittiest micro-controllers. C is already very niche at this point, and increasingly being replaced by C++ for embedded work.
BRB, cutting off my dick so I can join in the Rust fun.
>Rust has traction and hype
No it doesn't have any traction.
It has plenty of counter-hype thanks to the social justice warriors on its board.
I'll pass. It's worthless to me and the fact that Microsoft endorsed it is a sure bet that it's going to die.
Other languages have gorillion of libraries and other existing code that are not aware of bc. You either have to give up on those libraries or rewrite them.
Might as well just use Rust.
The C++ features are nice for greenfield projects, but they can't be cleanly integrated into legacy codebases without extensive refactoring. The defaults in C++ are also just plain wrong. Almost everything should be const by default. Heap objects should be wrapped by unique or shared ptrs. Functions should be nothrow by default.
It's too late to change these in C++ now, but at least Rust has gotten these right afaik.
leaving this here
64.github.io
C++ is bad for a lot of other reasons that no borrow checker or lifetime tracking will ever fix. It's just that it's the only popular option in its specific niches.
Anybody who tells you that they actually enjoy writing C++ is either a liar, hasn't been a programmer long enough, or is one of those myopic "career C++ programmers" who has never used any other language past a surface level.
>Even actual "safe" languages like Ada Spark require language extensions beyond what any C/C++ developer has probably dealt with to be fully verifiable.
That's fine. I can deal with that.
t. Aerospace engineer who writes verifiable code in C.
>Using C is a constant developer overhead
Only if you hire pajeet
> Hunting for bugs are runtime with C which would have been caught at compile time with Rust is overhead
That's literally impossible if you know what you're doing.
>Rust also protects against many concurrency bugs which C has zero support for preventing.
Sure there is a way of preventing it. By knowing what you're doing and through code reviews.
>want a memory safe C++ project
>add clang-tidy to the CI file with modernize-*, bugprone-*, cppcoreguidelines-*, and clang-analyzer-*
>use -Werror to force these changes to compile
>also add -fsanitize=address in debug builds and make sure tests pass
gee that was hard
and the clincher here is that static analysis tools are actually integrated with the language and its particular concrete use, and not some abstract cover-all idealism like integrating it hard and fast into the type system
>someone might accidentally kill someone with a knife so let's all just go ahead and ban all knifes and use cardboard in the shape of knifes to do cutting instead
That's literally your thought process.
>im not a rust shill
> rust has a cachy name, not a worn out reputation
>hype
The people who enjoy C++ are similar to those who enjoy Haskell. They love learning arcane trivia about their language and are predisposed to esoteric Buddhism.
>C++ is bad for a lot of other reasons
Works for me
>Anybody who tells you that they actually enjoy writing C++ is either a liar
You're full of shit
> hasn't been a programmer long enough
22 years and counting you little shit.
> myopic "career C++ programmers" who has never used any other language past a surface level
I've used more languages than you've probably heard of.
rust considers SIMD intrinsics "unsafe"
literally every game project is just a rust wrapper around a C or C++ DLL file with unsafe{} wrapped around every import
the absolute state of rust
If you know what your doing then why do you need code reviews? You never write bugs anyway, so why review anything?
>Someone told me that knives are safe if you know what you are doing, so I made a chair entirely out of knives that's perfectly safe if you sit in it just right.
That's literally your thought process.
>The C++ features are nice for greenfield projects, but they can't be cleanly integrated into legacy codebases without extensive refactoring
Encapsulate and then refactor.
> The defaults in C++ are also just plain wrong. Almost everything should be const by default
Matter of preference.
> Heap objects should be wrapped by unique or shared ptrs. Functions should be nothrow by default
That would cripple the power of the language.
>abstracting an idea to the point it doesn't even resemble the original concept and it doesn't work anymore
Yup, found the OOP codelet.
You know C/C++niles are threatened by Rust when they feel the need to respond to these threads so aggressively.
>If you know what your doing then why do you need code reviews
>If you know what you're doing you don't need peers to cross check your logic
>t. 18 year old zit faced Rust faggot
>increasingly being replaced by C++ for embedded work.
I don't see this at all. Size is critical on embedded devices, and naive C++ generates a ton of assembly that you have to pray that the compiler optimizes away. What's more, most embedded programmers aren't actually that great at programming, they're usually electrical engineers that have picked up C on the side and shit out some of the most badly-structured code you can imagine - trying to imagine them writing C++ just terrifies me, and they would never get a single thing done with Rust because they would never get past compiling "Hello World!"
Rust doesn't even support OOP. Cope and dilate cnile.
>Daily threads shitting on C++
>Daily threads hyping Rust
>All Rust does is complain that you're not using it and using C instead
LOL
Rust would take literally a day to compile some of our larger codebases written in C++. It's just not practical.