>go to Sup Forums
>tons of people shit on Rust
>they only make Sup Forums-tier comments about the community
>no real criticism about the actual programming language
>C programmers call Rust programmers brainlets even though Rust has many more features than C
Really activates my neurons.
Other urls found in this thread:
benchmarksgame.alioth.debian.org
ponylang.org
github.com
twitter.com
a
>>they only make Sup Forums-tier comments about the community
You must have missed all the technical criticism of Rust.
Cniles hate rust because its literally too hard for their simple minds. They lose their l33t points to it.
It has more features - those features consist of crutches, straight jackets, and other handicaps so Pajeet can pump out shitty code for a penny an hour.
more like
>Hey user, convert your PNG image to static webm. Your PNG is unsafe and triggering. Webm is also newer.
No but you can ask Hiro for WebP
JS has more features than the C too.
Is that even supposed to criticize anything?
It shows that OP's argument "proves too much", therefore fallacious.
I wasn't talking about which language is better in that context but which language requires more knowledge
>I wasn't talking about which language is better
I wasn't either.
Not OP but I certainly must have since I have never seen even one in all these threads.
>Sup Forums is bad meme
Common Rustfaggot tactic, blind to technical criticism
That's the reason C is just better. It's portable, minimal, and fast.
Rust has a lot of great features which it pays for by having a shitty slow obese compiler. Not to mention having to deal with the borrow checker every time
>Not to mention having to deal with the borrow checker every time
If you're having issues with the borrow checker then in C you are writing code that either segfaults or ignores const correctness.
Because it's pointless for someone who actually knows Rust to tell people who don't know any Rust at all about why it really sucks.
And it does, it does suck a lot.
Don't bother, most of the critics have no idea about the actual language and some of the so-called C programmers in these threads have never even been beyond fizzbuzz.
Pajeet won't be able to comprehend borrows and lifetimes and will rather wrap the entire function into if(thing != null)
>It's portable
Have you ever tried to make a real project that compiles both on Linux and MSVC Windows? Let alone runs.
>go to Sup Forums
>34 daily threads tell you to install Firefox
>1 thread shits on Rust
>49 threads pop up to defend Rust and SJWism
Really activates your Mozilla """outreach community manager evangelists""".
Good point, Steve. And Pajeet won't put up with your compiler rejecting his program, so he'll use unsafe code instead of spending a week trying to get his shit to work with safe code, making all your security benefits moot.
I'd trade one daily thread telling me to install brave for 34 daily threads telling me to install firefox desu.
i suppose that's an eternal cycle:
>user makes a thread asking how to get started with programming
>other anons tell him to start with C because "muh foundations" and that's how they started as well
>user studies C for a few days or months
>writes hello world, fizzbuzz and some algorithms
>user thinks C is the best thing ever because that's how he wrote his first toy programs
>user joins the hivemind of C evangelists on Sup Forums and never learn any other language other than C++
>It's portable,
Nivva please, the only thing portable about C is its syntax. That's right, not even its semantics are portable.
Rust has that too, but also has programs in it written that are portable.
>minimal
That's another word for 'lacking features as if it were golang'.
>and fast.
So is Rust.
This, but none of those was mentioned here.
>tries to force own non-standard language-specific package manager and build tools
instant drop
I hate nothing more than this shit.
Don't make me learn a bunch of new tools just because you were bored and wanted to make yet another make replacement.
>having to deal with the borrow checker every time
It's only a problem in the beginning, once you get it, you'll stop seeing the compiler as a thing to fight again, instead it begins to help you with sensible and concise error messages.
more like every programming related thread begins with shitting on rust as the first response or so
before we go any further in the discussion answer these 2 questions first:
>can i write the software of a digital-watch or a traditional calculator in rust?
>is rust faster than c?
if the answer to any of those 2 is no then dont even continue preaching about the shitty features of your language,because yes,sometimes features are bloat.both in terms of speed and in implementation complexity.because believe it or not when it comes to hardware programming its important that how fast can i write a working compiler for that new architecture.
pic rel perfectly presents your enormous ignorance
1st depends on chip used mostly, not all of them are supported right now
2nd depends on the situation
again, your ignorance shows no bounds whatsoever
There's nothing stopping you from reimplementing cargo using a makefile.
But I don't know why you would want to do that.
why are you avoiding answering the question my pink haired millenial friend?
>Really activates my neurons.
Kill your fucking self back to Sup Forums and take your unfunny underage memes with you kid.
But the answer was both complete and correct.
you are still not answering the question.
>is rust faster and uses less memory than c so i can use it to program microcontrollers...
>is a rust compiler easier to write for a completely new platform and architecture
or is this 2 question is too triggering for you brainlet?do you need to run to your safe-space when you see/hear these 2 questions because they didnt thaught you complex mathematics/optimization/memory management at woman's studies?stay as a webdog we dont want dumb,ignorant people like you at systems programming jobs.
Epic ownage bro you showed him
Hello well poisoners, yesterdays thread had tons of sound arguments against Rust, none of which you could refute and now you're here filling this thread with ad-hominems.
>can i write the software of a digital-watch or a traditional calculator in rust?
Depends on LLVM supporting the architecture. If the company producing the digital-watch also provides some sort of SDK and drivers it might be painful to actually develop for the device.
If you're actually curious about embedded development in Rust, look up the work of japaric and the RTFM system.
>is rust faster than c?
This is a dumb question, both compile to the same assembly, if you write Rust code equivalent to C you'll see the same speed, assuming you're using clang to compile your C. If you're writing higher level Rust, then the compiler actually has more room for optimizations, but then it also depends on LLVM being able to take advantage of strict aliasing, which it currently doesnt.
So is Rust faster than C? Depends. Is it slower? Also depends.
Only higher kinded types and varadic generics, they're valid criticism but you don't even use them.
Yes and yes.
Moreover, Rust semantics allow for optimizations not possible in C, so it's actually faster.
I was one of the persons giving sound arguments against Rust, and the only reply I got was "b-but they're just s-small deficiencies".
So it proved pointless to talk about what actually sucks about Rust.
>is rust faster and uses less memory than c so i can use it to program microcontrollers
It's as good if not better than C while also allowing you to write higher-level safe code.
>is a rust compiler easier to write for a completely new platform and architecture
Rust is using LLVM so you just need an LLVM backend.
>C programmers call Rust programmers brainlets even though Rust has many more features than C
as a C++ programmer, let me tell you, more features in your language does not make it better or "smarter"
>no real criticism about the actual programming language
that's because there is nothing wrong with it, and nothing great about it either, it's just average, prepare to see a lot more of these "it's exactly like C++ but with my favorite pet feature" type languages in the future
can i see some proofs to those claims?
That's actually false. It was true about 4 or 5 years ago, but everything changed after they wore the borrow checker.
>I was one of the persons giving sound arguments against Rust
Link to the post.
The only relatively sane discussion in that thread was with the honeywell guy and that was just "but you can statically verify C using our inhouse proprietary tools, why people refuse to write safe C code is beyond me".
Look at the benchmarks game or ripgrep.
Find me the archive and I'll link to the post, too lazy to look it up.
It is a language for trannies who have fried their minds with hormones
...
I can assure you the trannies wouldn't be able to actually use Rust.
The other thread, there were two threads about Rust yesterday.
This? Only 25 posts, no wonder I missed it.
Trannies can design processor architectures.
Why wouldn't they be able to use Rust?
except it's not, everything added since then is just a bunch of generic FP shit, every popular language (including C++) is putting these in now, it's nothing special
>Trannies can design processor architectures
Women will use this as proof that women can into technology.
If you would actually program literally anything that uses a library then you'd hate C/C++ for the rest of your life. Having a package manager like cargo is a bliss.
Rust is pretty much as fast as C. Besides, speed doesn't always matter as much as you think it does. Not to say that it never matters.
>m-muh bloat!
You don't pay for what you don't use.
Here's an example, trim a string in C and then split that string into words seperated by whitespace. In Rust that's just
let new_str = mystring.trim().split_whitespace();
If you wanted to do the same thing in C you would have to write a lot of boilerplate code and the result would be the same at best.
tfw C++ still can't split strings without messing with fucking streams
Nope, guess there were three threads.
I was the one who replied to you and actually I agree that they're bad things that should be improved upon. But having higher-kinded types or PartialEq on arrays is not critical in every context (I happen to work in a context where these features aren't critical).
Besides, the advanced points you were making are drowned in the sea of complaints about how borrow checking is for SJWs.
Rust is a great language, the issue is really that it's kind of limited in scope.
In a setting where a GC is acceptable, GC'ed languages make you much more productive overall, and in most cases you don't need Rust's amazing concurrency features. Rust is a great language for writing safe, deterministic and fast programs, but most programs don't have those requirements and something like F# would make you infinitely more productive.
For low level systems programming, C compiles to far more platforms, while Rust is currently hamstrung by only having an LLVM backend.
So the main issue with Rust is that its niche is fairly narrow.
strtok is cute! CUTE!
The work should be on adding more backends to LLVM rather than adding more backends to rustc, shouldn't it?
That's false.
They moved from a retarded smart-pointer design with special sigils, to a borrow checker with smart pointers in the type system which you rareely use and statically verified memory safety.
There are no other languages that have that set of features.
Find me another language with the same growing ecosystem, sum types, type classes, lifetime tracking, no garbage collector, no-overhead FFI.
I'm talking about the average tranny/SJW person, there are always exceptions.
Out of those only the lack of specialization affected me. I don't understand why they released 1.0 without that.
Higher kinded types would also be nice. I wouldn't be surprised if they started introducing breaking changes with the new epoch system they planned out.
Mind giving an example of a problem that negative reasoning solves?
>So is Rust.
benchmarksgame.alioth.debian.org
>In a setting where a GC is acceptable, GC'ed languages make you much more productive overall
That all depends on what language you know best, the borrow checker doesn't really limit you once you become productive with the language, that can be said about any tool really.
>and in most cases you don't need Rust's amazing concurrency features
Are we talking about the same language? Rust's concurrency stuff is awful compared to Go or Erlang.
>For low level systems programming, C compiles to far more platforms
That's because companies just patch up a 10 years old version of GCC for their hardware and give you some half-assed SDK.
>while Rust is currently hamstrung by only having an LLVM backend.
That's how LLVM works, you add backends to LLVM not to rustc.
It's the same as if your company made using clang mandatory.
>Find me another language with the same growing ecosystem, sum types, type classes, lifetime tracking, no garbage collector
C++, Dlang if you turn the gc off
>no-overhead FFI
please call me when rust has c++ abi compatibility, oh wait it won't because none of these so-called "c++ replacements" will ever have this
Erlang pays a high runtime price for its features, while Go's concurrency is inherently unsafe.
Rust's green threads got scrapped so it doesn't have as many concurrency options as I would have hoped, but the way the language is designed ensures that the language is safe for concurrency without paying a big runtime cost for that safety.
well, not even c++ has c++ abi compaibility
>benchmarksgame.alioth.debian.org
>n-body
>C version is hand-optimized with SSE instructions
At that point you are no longer comparing languages and compilers. Might as well use inline assembly.
If you want to use C++ template libraries from a language that isn't C++, Julia is pretty much the only option I know of.
Julia's language interop is pretty crazy.
>Sup Forums tier
Not an argument.
That's pretty good yeah.
>has programs in it written
bwahahahahaha
>both compile to the same assembly
just like go and pascal
>So is Rust faster than C? I wish it were. Is it slower? Yeah.
ftfy
C++ doesn't have type classes, it doesn't even have the gimped version that was proposed. It also doesn't have sum types, std::variant is awful to work with. It doesn't have lifetime tracking, it just has a couple smart pointers. All it has is its ecosystem.
>D
Hahahahahaha, you're baiting right?
>>That's how LLVM works, you add backends to LLVM not to rustc.
Ehhh, sort of. LLVM bytecode isn't portable, hence the "low-level" term.
With that said, Rustc should be fairly easy to write new backends for considering that it is now a three-stage compiler than the Rustc IR is both fully portable and fairly simple.
>just like go and pascal
Go has a pretty fat runtime and a garbage collector, they don't compile to the same assembly, I was being literal, not saying they compile to the same ISA.
Pascal would probably fit, but not sure, only used it briefly in school decades ago.
Only c parts are fast
It's still awful compared to Go or Erlang, the only nice things is that you don't get data races and the type system can prevent thread unsafe things from happening if used properly (see the MutexGuard issue that made it unsound).
I'd rather pick an Erlang if I needed concurrency, until Rust provides the required tools to implement green threads in a library I'll keep using Erlang (or Elixir).
Rust is going to be easier to write fast code in than C in practice, just like how C++ is easier to write fast code for. Just having proper generics is already a huge improvement. Having a ton of zero cost abstractions is even better.
C can be fast in theory. In practice, far too many things end up hidden behind a void pointer.
>Out of those only the lack of specialization affected me. I don't understand why they released 1.0 without that.
Because they haven't solved it yet, if they had delayed a stable release by years because of it the language would have failed. The same thing happened to macros.
>Higher kinded types would also be nice. I wouldn't be surprised if they started introducing breaking changes with the new epoch system they planned out.
I don't think they'll ever actually introduce higher kinded types, the most I expect is ATC, but still, gimped version.
>Mind giving an example of a problem that negative reasoning solves?
When coupled with specialization you can have a marker trait implemented for
some things that SHOULD NOT be specialized. That's just one example.
Another example would be specializing for the diamond problem, i.e. you want to specialize for `T: !A + !B`, `T: A + !B`, `T: !A + B` and `T: A + B`.
>but you don't even use them
Yes I fucking do you little shit.
Rust is also lacking CTFE, good platform support, and a non-shit borrow checker.
What's wrong with the borrow checker other than lexical lifetimes?
>C programmers call Rust programmers brainlets even though Rust has many more features than C
No, Rust is the real C successor. It's safer and faster than any other C competitor. Comfy type system. Easy cross-compile. Wonderful docs. Don't mix "Sup Forums brogrammers" with "C programmers".
Nonsense. Pajeet uses go, java and php.
More nonsense.
PNG could be pretty efficient already.
Too coarse.
>using MSVC
I think it's the right amount of coarse, with something more fine grained you end up with Pony, and you don't want to end up with Pony.
>Here's an example...
It means fuck all, you're using the standard library and not the capability of the language itself.
Let me see how do you implement a data structure or an algorithm. You'll need unsafe everywhere to be as fast as C.
>I think it's the right amount of coarse
Absolutely not. The compiler rejects perfectly correct programs, that is not okay.
>with something more fine grained you end up with Pony, and you don't want to end up with Pony.
What the fuck are you going on about?
A finer BC means you get less programs rejected.
not him, but I missed the memo stating that using standard libraries was considered harmful somehow
It can replace C for many tasks that C was misused for, but it doesn't replace C for many of its current applications.
It's much closer to being a C++ replacement imho.
>Absolutely not. The compiler rejects perfectly correct programs, that is not okay.
I know it does, but mind giving examples where it's actually possible to verify?
I've never actually hit a limitation in the borrow checker in practice.
>What the fuck are you going on about?
I'm talking about Pony, the programming language: ponylang.org
Rust's BC is the right amount of coarse. If it were any less restrictive, it would be impossible to formally check it. Frankly you could argue that it's *too* permissive in a sense, the type system is bleeding edge in many ways.
>Rust is pretty much as fast as C
"""pretty much""" ;^)
>You don't pay for what you don't use
"""zero cost memes"""
>Here's an example, trim a string in C and then split that string into words seperated by whitespace. In Rust that's just
>let new_str = mystring.trim().split_whitespace();
char *split(char *s)
{
while (*s == 32) s++;
return s;
}
the rust version is bloat
Why the fuck would a finer BC result in "ponylang"?
>growing ecosystem
>Dlang
dlang was dead on arrival and this was before a shift in the language compatibility that was comparable to python jumping from 2 to 3, the only people that seemed to use it were doing so behind closed doors
Look into it, you'll figure it out.
Now do it supporting Unicode spaces and UTF-8.
>Being this retarded.
Lmao. Don't you know that there's currently work going on to make Rust's BC finer? You don't even know what you're talking about.
github.com
Only a fucking retard would think Rust's BC is anything but too coarse.