>10000000 libraries, game engines, frameworks, etc. written in rust>0 softwareRustroons don't write code, they just partake in intellectual masturbation, since they only care about the language itself and not the software they write in it.
op here, whoops wrong image!
>>109389508Rust programmers are anti-semetic and racist. I do not like them.
>>109389508no, async closures are a bitch to work withalso, the borrow checker is not implemented as well as it could be: let array = stack.split_off(starting_index).into_boxed_slice(); stack.push(Value::Array(array)); I have to use the temp var arrayif I put it into stack.push(...) then the borrow checker complains
let array = stack.split_off(starting_index).into_boxed_slice(); stack.push(Value::Array(array));
>build times>micro-dependency hell>affine memelol
>>109389730you're actually fucking retarded if you don't appreciate microlibs.you're literally too lazy to fucking write the bit of extra boilerplate in exchange for more examinable programs.you also seem like the type of person to put "util" files into your projects, which would be better suited by microlibs.fucking retard.pic related is you (on the right at the desk)I'm the daddy
>>109389508
>>109389756>you're actually fucking retarded if you don't appreciate depending on a thousand libraries from strangers that could have vulnerabilities or breakages at any timeAh yes
>>109389508I like rust but I'm not gonna be baited into arguing for it anymore, i got shit to do.
>>109389788t. doesn't understand pinned version numberst. has never read the code of a lib he used (which is obviously easier when the lib is small)t. almost certainly cannot even describe what he means by a "vulnerability." perhaps he will attempt to do so in his next post?
>>109389508It's alright, but I generally prefer C for quick systems work.
Rust is coming to rape you in your sleep.
>>109389959I prefer C for literally nothing because it has no containers. Even Assembly has the stack.
>>109389508This.
>>109389970>no containersmalloc and free?Assembly's stack is just alloca in C and at that point you may as well use VLAs.
>>109389959>>109389970You guys needs to hop on Zig or you're nocoders.
>>109390050I tried Zig, it's too Rust-like for my tastes and kinda defeats the point of using C. If I were to use Zig, I'd just use Rust instead.
>>109390119Rust is>annoying to manage memory.>weird enums.>Takes forever to compiles (althought with some annoying setup it's manageable to shrink that).Zig has >good C interoperability while rust have to handle a day of work just to get a FFI working.Yes it does defeat the point of using C, but C is really annoying to use on the long run.Working with headerfiles? ffffffffuck no more.>builtin testing with fuzzing capabilities.>One of the nicest build systemI'm still using rust because rust has the features I specifically need and because I'm a retard that can't be trusted with memory management.But Zig is definitively nicer than C if you're actually spending a lot of time coding and not as a hobby.
>>109390213>long runYeah I'd use Rust. Like I said I just use C for quick work. The Rust enums are great and memory management is completely fine.
>>109389866You expect Normie's to actually read the code of libraries they import? The whole point of importing libraries is so that people dont have to worry about the underlying logic. Few people actually read that stuff.Instead they blindly depend on it. I don't know why you are acting like blindly depending on something without understanding it isnt or couldn't become a vulnerability. You made the point of microlibs being easier to read since there small but your forgetting that there is an extremely large number for them. Anyone who would be willing to go through all of that is probably better off making their own libraries.
>>109390270Someone will read the library.
>>109389508Just like (You), OP
>>109389543we really need more threads about this. Discussions would also be a lot more interesting than rust ragebait
>>109392129there aren't enough second/third hand talking points /g/eets can at least superficially grok and use for this to work.maybe if hk manage to trigger (((them))) somehow, it can become workable via (((e-celeb))) tard osmosis.
>>109392081Thank you!
>>109389866>t. has never read the code of a lib he used (which is obviously easier when the lib is small)who has time for this at work? nobody
OP haven't you heard? Rust is in its Extinguish phase now. There's no need to keep posting about it!
>>109392848To make that argument is to admit that:1. Rust libraries all have excellent documentation.2. You're a retard.3. Both.
>>109393149so for example, if I take the pic rel package, you really want me to not only read its code but also its dependencies code to be safe?>hey boss, sorry, can't work today, need to check millions lines of code first before I install that package.yeah GG
>>109393193there are zero dependencies in your picrel. i hope you're baiting. because you having sub-50 IQ* would be really sad.* IQ used colloquially as a substitute for an actually good intelligence measurer, which real IQ is not.
>>109389866I use rust, but realistically the main reason to use libs is to outsource the maintenance to someone else. If you use libs that are small enough to just read their entire code, and you do so, you might as well just implement the same thing yourself or vendor the code directly. If you're reviewing every update to every small lib anyway, you're doing all the code review work, while being subjected to upstream churn that you may not care about at all and no direct control on the actual code, so at that point it'd be easier to just make it your own code so you only make changes when you actually need to and review them internally.
>>109393262(nta)you should at least skim your direct dependencies retards. and this is not rust-specific.it's not like you're adding newfound dependencies everyday. in the rust ecosystem itself, you're usually dealing with dependencies in the lower hundreds TOTAL in ALL your projects and INCLUDING indirect deps. And some of those have public audits too with tooling that allows you to know about them.
>>109393193I want you to be able to identify what makes a lib trustable, which is its AUTHOR. assigning trust to specific projects is an incorrect method of operation, you assign trust to authorship.look at the RustCrypto projects---all of them are microlibs (you specifically import Argon2 for example), but all developed by the RustCrypto team, which is who you assign your trust to.by wrapping everything into an umbrella project, you increase compile times, potentially bloat binary size, and reduce visibility into what exactly functionality a project is using when looking at its dependencies.
>>109393193You'd have to do that no matter what language you're using. Idk what Rust has to do with this.
>>109389788Shows us the TLS, ssh and crypto libs you've rolled on your own
>>109389508It's the best available tool but it's far from perfect. Hopefully innovation continues in the age of LLM slop.
>>109390270You made a point of nobody reading what they import, but you're forgetting the fact that is also true for macrodeps. Slicing the same code in 2 or 2000 pieces doesn't change the amount of it. In practice however, any time you import more than one large macrodependency they will have a lot of overlapping code that does the same thing. Meanwhile, since microdeps are smaller they are more reusable, so two microdeps with large dep graphs of their own will not only have large overlap in their own graphs, but also between each other, actually reducing the total amount of code.
>>109393193Why are you under the impression that if instead of using the equivalent crate to import equivalence traits for key comparison hashbrown would have that code in itself then somehow the lines of code would magically decrease?
>>109393262If you use a dependency A which uses a tiny dependency B, that's small enough to just implement yourself, then when you need thatvfunctionaloty directly, if you reimplement B you end up with more code to check than if you'd just imported B
>>109389508>i need a compiler to enforce lifetimes because my brain isn't capable of itfaggots.
>>109394807Less thought overhead for competent people like me.
>>109394807>I want to perform work mentally that could be automateddo you even understand the purpose of programming?
>>109394832Define what do you mean by the "work".
nta, but...>>109394829>not thinking is what makes me competentyeah, sure, you solved double free but your code logic is still profoundly retarded.i think the root cause is just user error>>109394832the only thing clear here is that you dont understand the purpose of an interfaceimagine writing an abstraction that ends up being way more complicated than dealing with the underlying mechanism, directlyoh, wait, you dont need to imagine- thats rust!
>>109394832>>109394870>shitty interfaces(and also sepples, admittedly. hes our inbred cousin, we tend to pretend it doesnt exist)
>>109394870With Rust you can be lazy and obey the compiler and worry about nothing. Ctards are constantly juggling pointers in their heads.
>>109394909>Ctards>Juggling pointer in heads>Be lazyI'm pretty sure this guy doesn't even know what nostd is.
>>109394909cuz we like this shitor want to pretend were something we arenttheres no inbetweenyou on the other hand seem to hate the subject doealso i have tangible results from designing my own interfacestahts why i do c.picture ballmer. now instead of saying "developers..." its "interfaces..."thats og c-tards. the rest are poserswho may become og ctards down the line too, but if you ask a c user "ytf do you use that archaic residue of a language"the answer is gonna be "i dont mind its "archaic", i use it exactly because i design my own interfaces. and in this way, its not archaic, its fine-tuned towards being light"
>>109389707This is literally good code though. The temp var will be optimized away and doing it on one line would cause that line to be over 80 chars
>char _full_path[10000]>stpcpy(full_path, path)>dirent_buffer_size = 100000;Ctards are NOT sending their best good lord.
>>109394983>The temp var will be optimized away and doing it on one line would cause that line to be over 80 charsnigga idgaf. rustfmt automagically fixes all of that.
>>109395017its a wip, what did you expect?its like working with a rough block of marble, that gradually gets defines into a beatiful, anatomically correct sculpture of a juicy pussyrustrannies cannot comprehend the joy of creation...
>>109394969actual good bait, bravo
>>109395069>ill imply its bate, that'll show emsaid the dweeb for whom freeing after use is so complicated they need a whole framework that takes care of exactly thatis my reading correct?
>>109395017Keep allocating from heap, faggot.
>>109395141(the anon who wrote the code)it could have been cleaner thoughthe other fag 100% has experience in Cbecause things could have been made cleaner. and will beits still a wip, handwave value just to get the thing to work, to debug and deploy itactually, it was created ad hoc, and then shoved into my lib with a todo so i can "grep" for it if/when i want to things become seriousim in the middle of upgrading my stuff rn
>>109395404Don't be offended anon. Your code is ok. I insulting Rust Faggots.
>>109395404Yeah, but you allocated from the heap.
>>109395404>actually, let me lie more real quick
>>109395426yeah, i knowstillintellectual honestywe *are* better than the rust faggotsand what makes us better is the fact that we expect more of ourselvesits ironically shown in the languages we choose-we roll with 100% responsibility, no guardrailsthey'd rather be in a safeboxthis is telling. and even in a non judgemental way- were just cut from a different cloth...but when push comes to shove...yeah, then they deserve being reminded of that, i sayin a way, they fight against c, but they dont understand that c is merely the tail result of a mindsetthat of having control over ones slice of reality.c is merely an expression thereof, i think. if we were really to go into the weeds of thingsand thats what makes c programmers *just better*. even when they dont program in c, currently, for this specific project.
>>109389508correct.
>>109395532>todo is right there in the screen...>>109395502its an ad hoc."just fucking get it done kinda deal bc im bored to death with building the fucking abstractions"im still rebuilding everything from scratch so this has to be doneits just fucking painful, and i have a thing going on, so indulge mei have other things to fucking do, i need fucking money, not a fucking perfect lib, beyond a certain point
>>109389707this is the kind of retard who gets confused by C++ value categories and writes incorrect code. you can just tell.
>>109395596contye, im really frustratedi thought it was gonna be a quick in-and-out, but now im stuck between "doing things proprly"and"getting it fucking done"and even"im still a fucking noob, and thats why i evole my libs to begin with"but also alsothe overaching pressure of delivering something to get paidso yehfucking malloc the name, free it when i leave the context bc fuck itamdahl said so and who are you to argue with the guy?
>>109390270>You expect Normie's to actually read the code of libraries they import?I do all the time...I needed to replicate OpenSSL's evpkdf, I didn't know how it workedsaw https://crates.io/crates/evpkdf/0.2.0/code/src/lib.rsread itunderstood how evpkdf worked and added this as a dependency anyway because it's so fucking basic and I don't need to do shit.you're just retarded.
>>109394969>end[-1]what the fuck are you doing?
>>109395645>>You expect Normie's to actually read the code of libraries they import?>I do all the time...you kinda have to when yorue serious about thingsi salute you anondont care what lang you use, thats whats supposed to be donebut alsoyou can be doing free pror maybe not even freeso post a screenshot of your current workeveryone has something current going onand if they dont, theyre not programming, so they should shut the fuck up
>>109395017as >>109389970 said, this is what happens to a m'fer when you don't have containers or a type system that's remotely useful or expressive. Now people learn why C++, no matter how fucking awful, exists.
>>109394576Just trying to get a working understanding of curl is already maddening.
>>109395707Off by one.
>>109395707lol, just read the codei get end from stpcpymaybe thats the thing, you misread that as strcpy or somethingstpcpy returns the endptrits fucking erotically sexysex with stpcpyimpregnate stpcpylick stpcpy's feetshes a chibi tranime innocent girlthers no reason you shouldnt abuse stpcpy in the sexual sensethats literally what SHE has been DESIGNED foralso yeah. stpcpy(). youre welcome.
>>109395707also also alsowhen you index into thingsits not a uint or ulong youre usingyoure using an intand tahts just one of the things that make C what it is- complete and utter freedom
>>109395722>post a screenshot of your current workI haven't written a line of code since last year, because I instruct codex (and previously claude) to make things for me
>>109389508OP here, i'm trans btw
>>109395880yeah, but also:youre a fucking retard and you dont even know what youre talking abouthopefully youre at least getting paid for passing for a fool, on the dailybecause holy shit are you a whorewho just got dumpedsofuckinghardor maybeare you doing that for free?thats would have been hilarioushonestlyhopefullyyoure not doing it for free, are you?
im f5'ing with atention, """anon"""im really curious what your marketing division is gonna come up withyou dumb fucking retarded whoreyeah>SHOMEyou fucking whore>SHOMEyou fucking piece of retarded shitai whatagain?
ai what fucking nothing, das whatjust like we fucking said 2 years agoso now, shoosh, you fucking mongoloid inbred retardsits a fait accompli nowyou can cope and seethe nowi fukken allow ithow bout that?
>how bout that>not even thankful that theyre allowed to cope and seetheungrates.legit fucking ungrates taht we tolerate them, only that they become less fucktardedwhat to do with em?we show mercy, and in return they do everything to destroy uswhat to do with them?time and again-inultinjurymurderhow much more can we take?
>>109395645> I read libs that means your wrong.Good to know that you speak for all Normie's you fucking retard.>>109394629>micro deps will have large overlap. This requires someone actually understanding the dependencies in order to make them overlap. While one large library wouldnt require this step.And most projects with micro libs hardly have any overlap at all. You act like importing micro libs garuntees overlap and it couldn't be further from the truth.>>109390302>B-B-But someone will read the library.News flash dumbass, that wasnt my fucking point.
>>109389518most of rust is happening in the proprietary world, just like C and C++ but both ahev 4 decades of foss behind them when rust only 1
>>109396219rust is a straight up mistake.its an errorall its gonna take to make that a status quo,is one side project
>>109396304it just works and fagman are using it, everything else is noise including the troon psyop, they don't exist in real life, only inside your head
>>109396349>open your head == downgrade and learn a ton of shit because ***OTHERS*** suck at basic programming things like freeing after usestill doesnt solve for fuckin retardsin factrust only accentuates the problemand im not concerned anywaysc is easy for me, idk where the fuck youre coming fromi use c EXACTLY because its easy>picrelyeah, no matter the frameworkretards are gonna retardand rust is creating flse positives>it compiles so its correctyeah no, youre mentally retardedsomeone pushed in on your fontanelle when you were a childand so you use rust nowdas the storyjust use arenas you fucking retard
>>109396490i'm not going to read your post as i don't care, regarding your screenshot.>muh rust cve>look inside>first issue is actualy C bindings related issuekek
>>109394969As a C++ codelet it always feels so weird that i can read C. It's not quite the same yet i can still follow along.Also i don't know why C++ dropped the convention of naming functions with '_fn_', i really like that.
>>109396516>you dont care so much you repliedrepeat after me:>i am>mentally>retardedits gonna free you from alot of responsibilities, you should try itbecause you truly are mentally deficientlikeim 37i know my limitsi cant program for 14 ghours straight like o used tobut holy fuck are you midbrokenyou remind me that despite being at the bottom of my performancei am still largely superior to at least 90% of humanityESPECIALLY YOU INCLUDEDYOU DUMB FUCKING RETARD SUBHUMANim drunk btwi fucking annihilate you, effortlessly, when im at the bottom of my capabilities, despite being an old fuck. and drunk.let that fukken sink in, faggot
>>109396490I do agree that Rust gives a false sense of security, because i always see people talking about it as if it's logic-safe or that Unsafe Rust doesn't exist.
>>109396551>i can't program for 14 hours straight like i used toEh, you probably get the same amount of work done if not even more in less time due to the accumulated experience.
>>109389508>println!kill yourself
>>109396545c++ is supposed to be a superset of c, so its only normal that you can read itthe difference is:can you work without all the fancy interfaces c++ provides you?at the point im at- learning c++ would be counter productive, because i can express all i want with cin fact, c even has name-mangling, so basically the same as c++'s zero-cost abstraction afaiki dont use it though. i dont use c to hide stuff from the user (most of the time-myself)c is really an interface thing. if you want these pre-made for you, you go c++, basicallyno judgy? question of prioritiesi do numbercrunching, so i need to control everything, so i have no use for c++its nto a matter of elitism, but of design constraintsand ive been doing that since 7 years now, so ive learned a thing or two about c, i have gotten used to c etctheres a dimension where both c++ and c are valable, depending on your usecasei still hate c++ but its because i can afford to, so to speak
>>109396579thanks for the positive vibesbut i wish the body could follow>>109396564people think they want a silver bulletbut what theyre actually wanting, is just to get betteri think learning another framework doesnt help with thatwhat helps is learning the fundamentals, but for reals this time, i think
>>109396516Many such cases. Just look at how Linux had 432 CVEs in one day as a counterexample. C is a shit language for retards like this faggot shitting up the thread.
>>109396667>c is a shit language because i cant deal with it>just look at what other poeople do with itim not "other people" thoughand im CERTAINLY not you, lolle
>>109396682It's a shit language because no one can. AI is proof of this. You can cope all you want, but it's over.
OP, now it's perfect.
>>109396691no one can except those who doyour firdge runs on c.airbus runs on c.satellites run on c.only your program doesnt.who's to blame, you think?c is not hard, its just taht your brain needs to bend in the right directionits something that develops with time, so you cant tell whether yloure bad at c from working with it for 15 secondsbut there is indeed a part of understandingbut but theres still a 50% copmponent of just being compatible with the frameworkthis isnt a question whether youre good at c actuallyor whether c is a good or bad languageits a question of pure growing up and understanding taht your not god, youre not able to do everything efficientlyand thats ok, because were all differentyoure only salty because you want c bc its a badge of honor for youwhile you can be much more efficient just going along the flowyou fucking nigger retard egotist
>>109389508Syntax is awful and I just don't like itI would prefer to read .NET code over Rust code. That's how bad Rust code is to read.
>>109396744And all those things are shit full of bugs and issues that pass through because of extensive QA, at best. Shut the fuck up retard. Holy shit. You retards are singlehandedly responsible for setting computing back for over 4 decades. I'm glad Rust is finally trying to reclaim some of it back from you fucktards with basic 80s tier thinking like sum types, monadic containers and lifetimes encoded in the language.
>>109396765nobody cares about this shit you academic jackoff. The only legitimate edges Rust has over anything else is the technical impressiveness of the borrow checker and memory safety.
>>109396777>No one cares about having language with a usable type system and actual movable types unlike my shit language that doesn't even have references. Shut the fuck up.
>>109396765yeah but noyoure just a salty idiot whose whole identity can be summarized in one word: "envy".just it gud, niggertheres at least one thing yorue good at, no?and tahts the direction youre supposed to go twardsyes, i can do everyrthing.but also youre a webshitter retard who gets paid 100k for 2h of legit work per dayyou cant have it alland thats your fucking problemyou want it all regardless whether you even deserve itand das not how reality works, faggotits an either/or porposition, on which you have no sayyou just roll the fukken dice
>>109396794>Usable type systemLike TypeScript? I actually do like TypeScript. TypeScript has an elegant syntax. Rust does not.
>>109396800Rust syntax is better than typescript.
>>109396820This. Imagine claiming a language that has insane shit like ASI has "good syntax." or had to add new binding declarations because var hoisting is weird as fuck.
Python matters until performance and therefore Rust matter.
>>109396820If you like Rust's syntax over TypeScript, when the majority of developers are already familiar with JavaScript and can easily pick up TS, I think you're a little bit crazy but whatever, it's your preference
>>109396859And you don't know shit about programming if you think the shitheap that is JS is good syntax. Write a fucking parser for it if you're so great (you can't). Writing a Rust parser is not that hard in comparison and you're delulu if you claim otherwise.
>>109396551cool b8 not reading all that though.
>>109389518Typst is written in Rust.
>>109396908>noooo dont share random shit/vent on 4chanseethe harder, its already donebut yeahdont feel obligated to read it, obviously
>>109396978>>109396908
>>109397056>>109397051>>109396908what the fuck is even a typist, lads?let me explain my position:>you can derive correctness from c syntax>then why do i need to deal with some retarded conventions to make the compiler's job easier?ah yeah, rightthe rust devs can do better, given rust is FUCKING OLD ENOUGH TO DRINK BY NOW? ISNT IT?FOR FUCKING COMPARISON:RUST IS 1/3 AS OLD AS CRUST IS 1/2 AS OLD AS C++AND ITS STILL UNSTABLEyeah nah, you tried, but your solutions obviously dont workand i dont mean the language-i mean the whole approach, actually
>the rust devs can do better*the rust devs can'T do better
>>109397079go to bed you are drunk
>>109397112thats not a counter-argument.i appreciate the concern, though
>>109397079C and C++ are unstable as well. They find new things to add to both and the compilers are all racing each other to find the next 5% performance boost by language lawyering the shit out of it. You're just a noob, because you don't actually do anything serious. Daily reminder C is finally getting basic shit like defer (already if you count the major compilers having some extension).C finally got checked arithmetic functions for over/underflowC had to literally steal C++ memory model because C tards think computers are still single threaded and hand to scramble to cope with the reality of symmetric multiprocessing. Oh and you no longer need xxd to generate c code to embed arbitrary binary data since CPP now defines #embed. Congrats cucks.You guys are just dumb as fuck.
>>109397125>thats not a counter-argument.neither is what you said, in fact you didn't make any argument whatsoever that's why i resort to trolling.make a clearly defined argument of why you think it's bad etc or don't expect a proper response.i like rust, i don't think it's perfect by any means but is sure as hell is better than c/c++, and i've used those over a decade before fully switching to rust, i'm just not going back as i'm having a better time building anything with rust than i ever did with c/c++.
>>109397128thats not the same type of instabilityc is backwards compatiblec++ can go fuck itself though, its literally c but when youre in a padded room>C finally got checked arithmetic functions for over/underflow>C had to literally steal C++ memory model because C tards think computers are still single threaded and hand to scramble to cope with the reality of symmetric multiprocessing.literally: what the fuck?all you know about c is what you read in articlesif anybody is erious about c in 2026, they write for an environment, and for a target configurationthats the thing about c, dummyand the reason why one wants to control their own interfacesbecause of all the shortcutsand because of the calling convention of your piece of code>nuh but your libsnoody fucking uses themand i bet my forskin at least 50% of users still work in the subset of C 99because we just dont fucking care, mangif we wanted bells and whistels, we would have been using c++well, c99 + compiler extensions of course,to each their ownfor me its gcc bc they enfore force inlines...but whatdya fukken know about it lamaoyou just want a python, but fast
>>109397190You are really aggressively projecting right now, just saying.
>>109397190>SEE IS BACKWARDS COMPATIBLE >real world C projects after upgrading gcc and clang: 100 new warnings!
>>109397190>you just want a python, but fastnta, yes but not exactly, i don't have to bother myself with complexity when it's not needed, but if now i want to fidle with some raw bytes in memory or some inline assembly or lower level stuff / whatever, i can, that's why i like rust.it doesn't get in the way of either building something, making an abstraction or getting my hands dirty when needs be.
>>109397199i mean he's drunk so there is that.
>>109397206Rust also has a real freestanding mode with no_std and no_alloc plus naked asm, so you can basically do whatever insane shit. Reminder that even asm isn't standardized in C and every compiler will do it differently.
>>109397190>>109397201>if anybody is erious about c in 2026, they write for an environment, and for a target configurationactually- *seriousalsothe standard != extended cits actually a subsetand undefine means "do whatever"but that "whatever" can be defined by the implementation (and it actually more often than not- is)>>109397199i accept your constipation
>>109397213I get intoxicated and act stupid on the internet too sometimes.
>>109397231I see you.
>>109397235so do i except my thing is more acid than alcohol.and generaly then i'd not argue about programming.
>>109397206yeah but as in all things engineeringits a tradeoffbw complexity/ease of use, power and convenienceor something of the sort, but you get the gisti use c mainly bc im used to itbut i got used to it because a function call is a meaningful portion of my runtimeand theres other uses for c so i will oppose your idea that rust is the silver bullet for everything and ever forevermoreand, in facti do appreciate its potentialand the philosophy behind itmy actual critique is that its uncookedand that not enough attention has been paid to ergonomics.
>>109397231>i accept your constipationlmao thanks for the laugh
>>109397260For me it's meth (sometimes) or weed or mushrooms or ecstasy or xanax or kolonopinIt depends
>>109397267
>>109397261>its a tradeoffi agree.>i use c mainly bc im used to itwell, that used to be true for me too, but now i'm more used to rust than anything else and i'm just comfy there.i always had some paint points with c ie the preprocessor, no standardized build system, package manager etc.>and theres other uses for c so i will oppose your idea that rust is the silver bullet for everything and ever forevermorethere are few niche things in which i still use c, but rust has replaced 99% of my use of it.and technicaly there isn't a single thing you could do in c that couldn't be done in rust, i generaly use c when i want to interact with something that has a c ecosystem and i don't feel like writting bindings / rewriting the whole stack.>my actual critique is that its uncookedi think C is as well, just look at the preprocessor.i respect c for what it represents, but i honestly think it's kind of an outdated tool.>and that not enough attention has been paid to ergonomics.i do have a few things that bugs me here and there with rust, but i find its ergonomics to be generally comfy, derive macros are also amazing.i like its trait system and what the metaprogramming features allow, some libraries could also simply not exist without those built in metaprogramming features, ie serde.
>>109397329I don't think that there is a replacement for C in some things or a pressing need to do so.C goes down into assembly and then you'd be reading ISA documentation, platform architecture documentation, it's all heavy and very very complicated, despite its deceptive "simplicity", it's actually quite complex. Abstractions serve a very valuable purpose. They enable us to not have to think about such complicated things. They are a relief.
At work my team had a choice of two different tools that did the same jobOne was proudly labeled as made with rust. Guess which tool we don't use.
>>109397372The one that didn't meet your needs?
>>109397261I just don't understand why the Rust syntax has to be so goddang awful. It's some like weird combination of Python and eye cancer.
>>109397329>i think C is as well, just look at the preprocessor.myeah but also look at modern cand the configurability of inlinessure, strictly speaking, strict portable c kinda fucking sucksbuuuuut... thats not c in 2026c in 2026 is c + the compilerand then it becomes truly a macro assemblerit really depends on the mindbut if you like to visualize the 1's and 0's individuallyc is your firendand any degree of abstraction above c is the enemybut then one tends to gravitate towards certain problems, and thats how one ends up working with cits a synergy of innate traits and work applicationsand so yeah, the preprocessor "sucks", but using modern prorgamming techniques in c makes it a non-issue, and actually, even a stylistic error, because the tooling doesnt interact very well with macros.so in modern c id say macros should be used for purely syntactic reasons like injecting parameters or abstracting messy formulations>outdated toolfor what you dobut when youre optimizing a hot loop to death and back, you have the choice:either you go asm, but then your dev time goes x10, if not better, depending on your aptitudesor you go with c bc youre ok with getting 99% of the performance with comparatively 10% of the effortand why c and not rust?bc at that level, c is just more convenient.i would have to learn rust to basically write c in it. its pointless.>the trait systemi theorized some aspects of teh trait systemi think you dont need an external syntax to c or a constraining framework to provide similar or even identical guaranteesjust think about it:the corectness of a program can be discerned from c syntax alonethat means all the information thats neeed is contained within the c syntaxthen why the fuck is there asomething like the borrow checker?just look at the fucking code, matec is perfectly cooked.its just that its consumed in a way you dont realize.it larps as a high level language but its philosophy is that its a macro asm
>>109397353>C goes down into assembly and then you'd be reading ISA documentation, platform architecture documentationHuh? I thought microcode was proprietary classified information.
>>109397425at best? autism.at worst: to be a designed cognitive burden to force people into big teams if they want to do anything funky.like re-creating blackrock's alladin, because even originally, in 1995 iirc it ran on a pc, and was written in c++so imagine what you could do with a gpu in tyotl 2026
>>109397439Who gives a fuck Mike, because I don't.
>>109397429>it larps as a high level language but its philosophy is that its a macro asmWhat i always understood C as is that you have the option of gravitating towards either mid-level (high-level back in the old days) or low-level, and that this is why C got adopted so quickly and became so widespread. It's very powerful.I can even see this in my Frankenstein C++ that gives this option as well (to a lesser extent).
C sucks. Who wants to use a language where containers introduce bugs?
>>109397448>at worst: to be a designed cognitive burden to force people into big teams if they want to do anything funkyI had never even imagined this, and the thought that people could deliberately design langs like this leaves me panick-stricken.
>>109397487Well i do because i'd like to read and learn them lol.
>>109397511https://nand2mario.github.io/posts/2026/z386/Here you go. 386 microcode full reversal.
>>109397429>either you go asm, but then your dev time goes x10, if not better, or you go with c bc youre ok with getting 99% of the performance with comparatively 10% of the effortnope, c won't magicaly result in more performant code than rust.you can write rust that compiles to the exact same asm as c code, there is no need to write C for that.in fact rust code is often more performant than c for a few reasons:1. it's less of a headache to use multithreading, gpu etc.2. since the language is more expressive you can give more hints to the compiler about what you want to do, it can then optimize better, just one example of the is function chaining iterators compiling to proper simd code.3. it's easier to pass messages to the compiler like "compile this function both normaly and with avx, ship both in the executable and decide at runtime which to use".>i would have to learn rust to basically write c in it. its pointless.i agree that rust takes a bit of time to learn and master, though if you master rust trust me that you will be a lot more productive than you ever were in c, the tooling alone make it worth it let alone the rest of the language features.>i think you dont need an external syntax to c or a constraining framework to provide similar or even identical guaranteesit's not about guarantees, it's about reusability, i'm gonna give you one example#[derive(serde:Serialize, serde::Deserialize)]struct foo {}now your struct can automaticaly be converted to and from any format known to man, third party libraries (for different format, ie toml, json etc) just need the struct to implement the trait Serialize / Deserialize (which was autoimplemented with a macro (only possible thanks to good metaprogramming)).traits impl being separate from definition means you can impl third party library traits on your own types (which mean they can now use them seamlessly) and your traits to library types, allowing you to extend them without editing their source
>>109397490yeah, thats the thing with c- it gives you optionswhat it doesnt give you, however, is guardrailsyou can go really deep down like working with avx stuff and then there youre basically dealing with asm directly, with the c syntax just doing the piping between thingsyou could use the libc for everything and then youre using a quite high level languagebut yeah, the tradeoff- bc theres always one- is that the framework trusts the user for correctness>I can even see this in my Frankenstein C++ that gives this option as well (to a lesser extent).sure and thats what is the "sane" approach to c, i hearlearn c++, use the high level features that you like,go down to c when neededi do purely c bc i did so much of it that im past the point of "why bother, i can even emulate portions of c++ if needed"i mostly think about name mangling. thats a pretty powerful featurebut also self modifying code, but then all bets are off, lol
>>109397504its not a certitude, if it makes things betterbut i think it is indeed a possibilitylike i said, conceptually, one should be able to derive most if not all the info needed from a c syntaxwhy hasnt it been done?>>109397541>nope, c won't magicaly result in more performant code than rust.never said thati said this:>i would have to learn rust to basically write c in it. its pointless.because this:>1. it's less of a headache to use multithreading, gpu etc.doesnt come free>more expressivethan vanilla cbut then the compilers come in, and then id hasard that its the other way around>serdeits just a matter of conventionc doesnt provide you with these because its supposed to be leanits not like its something hard thoughslab + offset and here you goyes, you have to write it yourselfand that kinda sucks, but thats the tradeoff to have a lean language
>>109397527Wow, i honestly had no idea that was even out there. I thought they had that locked in a CIA vaulted office or something lmao.You have my gratitude man.
>>109389508Perfect for AI.
>>109389970>Even Assembly has the stackYou know your can just declare a local array in C and get access to that same stack, right?
>>109397585> would have to learn rust to basically write c in ityou'd not be writting C in it, and if you think learning a new thing is pointless then i'm sorry anon.i think it'd make you more productive and have a better time once you are good at it altough the learning phase may be frustrating.>doesnt come freeliteraly does.>but then the compilers come in, and then id hasard that its the other way aroundnope>c doesnt provide you with these because its supposed to be leanno, it doesn't because it's not capable to.>slab + offset and here you goi think you don't understand the featureset of serde, you can do per argument flags to skip, change capitalization on the format file etc.but mostly, the thing is that third party library can just convert your struct to any format, they just rely on it to impl Serialize / Deserialize.and serde is just one example, tons of other libraries are just as nice ie Clap for parsing cli commands.and nice thing is you can mix and match things.>yes, you have to write it yourselfand that kinda sucks, but thats the tradeoff to have a lean languagewell, there goes my productivity argument, yes C is a leaner language, but how does that help you when you have to write 10x more code for the same thing (and it probably won't peform as well)?think about it, time you'd spend writting boilerplate is time you could instead spend optimizing hot paths.also we go amazing tooling for profiling, ie flamegraph, tracing etc
>>109397699>and if you think learning a new thing is pointless then i'm sorry anon.because learning shit just for the sake of it has any value?plus: its not like im *entirely* rust-analphabete, its actually not that hard to understand, most of itbut its orthogonal to how hard is it to *read, or to work with>>but then the compilers come in, and then id hasard that its the other way around>nopeuh-huh>i think you don't understand the featureset of serdeserde is serialization, deserializationeverything else is extraim not sure what youre thinking about>well, there goes my productivity argumenthavent you ever heard of modularity/code reuse?cmon, anonyou can do better than that
ackshuallytheres a profoundly zen thing i have to say about>>and if you think learning a new thing is pointless then i'm sorry anon.>because learning shit just for the sake of it has any value?we are limited in matter, because were incarnatedour forgetfulness is an optimization technique to make the best use of the hardware we have at our disposaland id much rather focus on the problem at hand than to be unable to ancompass it because i spent my resources on knowing all of the alternatives which have no use right now
>>109397798*encompass italso yeahthats why i like cit doesnt get in the wayand i can entirely focus on the problem at hand>b-but the double freesyou seriously think its a problem?we operate on different levels then
>>109397768>because learning shit just for the sake of it has any value?yes, but there isn't any value, i'm telling you rust is a much better language than c and not only you will be more productive with it, you will have a better time.in the end you do what you want, but me and all the friends i know that switched to rust don't regret it and aren't comming back.>because learning shit just for the sake of it has any value?that's just an habbit thing, i find rust to be much more readable than C is now, in part because it doesn't need to be as verbose.you can have 100 sloc of C turned into 2 lines of rust function chaining in many cases.>uh-huhcargo gives you full control over the compiler's args / toolchain etc, there is literaly not a thing that you can do in c you cannot do in rust as it was one of the goals it was built with.>serde is serialization, deserializationi was just giving you an example of the kind of libraries that are possible in rust but not C, there is a lot more to this, you are kind of missing the forest for the tree here.>havent you ever heard of modularity/code reuse?yes, my point is that rust allows more modularity and code reuse than c does because of metaprogramming, traits etc.god C doesn't even have the most basic things like templates.
>>109389508>no break n/continue n>no non-local control flow>no translucent adjunctive temporal scoping
>>109397798the thing is C DOES get in the way, you just don't see it because it's the only thing you know.we can make a bet if you want, write some whatever application for whatever spec (without llm use) and you'd always have your thing done faster in rust and it'll be more polished and perform better to.> than to be unable to ancompass it because i spent my resources on knowing all of the alternatives which have no use right nowthat's just an excuse to not learn something new, you don't have to learn ALL the alternatives i'm literally telling you which, i've learned a LOT of technologies, and rust is currently one of the only ones worth learning if you are a C guy.>you seriously think its a problem?>we operate on different levels thenuse after frees and double free can both lead to arbitrary code execution, but that's not the point, you mentioned that out of nowhere, at no point was i even arguing about le "muh safety" i'm just telling you rust is a much more productive language to build anything with than c, that's just a fact, whether you like it or not is not my problem realy.i'm not using rust for le safety, i'm using it because it's a good language that allow me to get shit done, the safety is just the cherry on top, honeslty i care more about the type system being superb than the borrow checker safety features, altough it's still neat.
>>109397810>i'm telling you rust is a much better language than cyeah, but you ignore the whole concept of usecase to say thator you dont even know thats a thinghow can i give credence to what you say?first the fact you dont know what a library is, and now you ignore the idea of use caseits not without reason that c is widespread after 50+ yearsforth diedlisp diedcobol diedbasic diedpascal diedc is still going strongand in new langs like cuda or opencl too>i was just giving you an example of the kind of libraries that are possible in rust but not Cyeah, i dont think you even program beyond this point.typical crab evangelist
>>109397832>>no break n/continue nboth break and continue are literaly rust keywords you don't know what you are saying>>no non-local control flowsecond skill issue.>>no translucent adjunctive temporal scoping>i'm just gonna write words i don't understand the meaning of now
>>109397832No need for break n when you have labels.
>>109396938No clue what that is
>>109397079>OLD ENOUGH TO DRINK BY NOWNow, I know minors can drink with parental permission, but if I had an 11 year old child, I don't think I'd be letting them have any alcohol.
>>109397860its 20. soon 21the rust project started in 2006yeah how time flies and how its impractical or something
>>109397841i wrote an elaborate response and accidentaly closed the tab so i'm not rewritting all that.>first the fact you dont know what a library iswhat makes you think that?everyone knows what a library is...>usecasei don't, name one, rust is a better pick in 99% of cases for new projects (if technical merit is your only reason, there could be non technical reasons, ie your coworkers, having to interact with legacy codebases etc).also who do you think i am, i used C/C++ for over a decade before switching to rust, i'm not some retarded zoomer. yes usecase matter to pick a technology, but sometime a technology is so shit that it's rarely the ideal pick.>irrelevant language listhalf of them are still going>its not without reason that c is widespread after 50+ years>first so must be best fallacythe reason is that it was one of the first higher level language still allowing you to do low level stuff, the preprocessor was a first naive attempt at it, however it sucks balls and we learnt from our mistakes to make better things.>c is still going strong>muh old so must be good fallacyphp is also still going strong, it doesn't mean that it is good, the reason is that there are still tons of legacy codebase and it still technicaly works (even if bug riddled).>yeah, i dont think you even program beyond this point.do you ?it is a fact that C lacks proper metaprogramming and thus it also is a fact that you cannot build libraries relying on it.also enjoy your 20 different build systems that all sucks>typical crab evangelisttypical boomer that doesn't want to learn anything new.your argument comes from a lack of knowledge on one of the two mine comes from deep knowledge of the two, you are basicaly in no position to argue because you do not know rust and thus do not understand what it brings.
>>109397857>i don't know so it doesn't existretard.
>>109390213>Working with headerfiles? ffffffffuck no more.What difficulties emerge with header files?
>>109390213>compile timemost of the time is actualy spent linking, change your linker to mold or wild and see your build time shrink by half.even on big projects, if they are well separated into crates, an incremental build is rarely more than 0.2s on my machine.
>>109397964>What difficulties emerge with header files?it's not that it's hard, it's just that it's fucking annoying and a waste of time / boilerplate.in rust you define your function once, that's it.no import retardedness, no macro pragma once / ifndef lunacy, just define your function and use it.
>>109397841>lisp diedin your dreams, mate. The community is alive and creating more libraries and stuff. The main compilers are getting releases and improvements all the time.>c is still going strongindeed and will continue to be used
>>109397844>both break and continue are literaly rust keywords you don't know what you are sayingreading comprehension>second skill issue.You chose the language with a compiler that babies you, not me.>>i'm just gonna write words i don't understand the meaning of now>I don't understand things people smarter than me do... WAAAHHHHHHH WAAAHHHHHHH MOMMY HELP MEEEE
>>109397985header files are optional. you don't need to use them if you find it tedious.
>>109398003>reading comprehensioncope, you just failed to write what you intended to mean.>You chose the language with a compiler that babies you, not me.it only babies you if you don't know what you are doing as cniles have proven to over the years.if you are good at rust you almost never have the compiler complain at something because you know what it'll be made about before you even save your file.>people smarter than me doyou sure think you are smart, go on, ignorance is bliss.
>>109398015>you don't need to use them if you find it tedious.that's false, especialy when writting libraries or when writting any sizeable piece of software.maybe for the helloworld.c you don't need them, but if that's what you are arguing for i seriously doubt your programming background.either way with C you are gonna have to duplicate code.
>>109397956>>first the fact you dont know what a library is>what makes you think that?that you pretty clearly implied that you rewrite all the boilerplate all the time with c.>>irrelevant language list>half of them are still goingnehyou cant say cobol or basic is still a thingyeah finance works on cobol but that bc theyre crusty old fucks>first so must be best fallacyit wasnt first, nor is the best full stop*its the best we gotand it constantly evolved since then>php is also still going strong, it doesn't mean that it is goodits comfy, have you used it? basically paiton>yeah, i dont think you even program beyond this point.do you ?yupwant some handles, kid?>muh meta-programmingthe meta what?youre programming in c, not inventing genders, son>20 different build systems>quiet admission that, indeed, crab doesnt write original codei use one. i dont care about the 19 othersits makefile. the most classic of them all.
>>109398020>cope, you just failed to write what you intended to meanI wrote "break n" and "continue n", and that's what I meant.>>if you are good at rust you almost never have the compiler complain at something because you know what it'll be made about before you even save your file.>If you are good at rust you get raped into writing code for the compiler rather than the machine and this is good because... BECAUSE IT JUST IS, OKAY?!>you sure think you are smart, go on, ignorance is bliss.I am smart, that's why I make $630k salary (tc 840) in hft.
>>109398038> i seriously doubt your programming background.is this you? >>109389970
>>109398052>that you pretty clearly implied that you rewrite all the boilerplate all the time with c.there is boilerplate you can't avoid writting because you are doing C.don't go and pretend that not half of c project don't implement their own shitty version of a hashmap for their types because c lacks basic templates and the preprocessor is a huge cope.and that's not even touching metaprogramming yet.:1a rust proc macro is a function that takes a tokenstream (basicaly the ast) as input, edit it with actual rust code and returns the ast.you can do ANYTHING in a proc macro.one example of how it can be used is in sqlx, you can do your sql queries, and at build time the compiler will query your databse to check for types mismatch between your queries, the types in your code and the actual database schemas, of course it's an optional feature disabled by default and i tend not to use it because it's a bit too woo woo for me, but that's just to give you an idea of what metaprogramming can do, no build time behavior / code processing is off the table and it's built into the language and thus also easily shippable into libraries.>you cant say cobol or basic is still a thingcobol is still massively used in banks.basic less so but i'm sure there is some legacy shit that still relies on it>its comfy, have you used it? basically paitoni've used it in one of my first jobs and honestly i don't realy like it, i do get it has improved a lot since but still, not my cup of tea, also a lot of jeets use it so that means you'll often have to work with terrible codebases.>want some handles, kid?i've probably been an engineer longer than you have lol>the meta what?>argument from ignorancemetaprogramming is basicaly writting code parse / write / edit code, but instead of a shitty script outside of your project, it's built into your language and can be part of libraries.see :1
>>109398052>i use one. i dont care about the 19 othersyes except every project uses a different one, there is no standard and it's fucking annoying.besides using libraries in C is so much of a pita that half the time you write your own shit.project tries to use as few libraries as possible because it's a pain in the ass to add them / rely on external code.you rarely see a C library that depends on a bunch of other libraries, not to the extent rust crates do, and that means you are sleeping on free performance.>its makefile. the most classic of them all.i've used make a ton, still a shitty ass option compared to what cargo can do.kek with cargo you can even link to a separate project (ie git repository) to use a single function of that project seamlessly if you want, not that it is a good idea but you can.
>>109398053>I wrote "break n" and "continue n", and that's what I meant.you seem to lack the reading comprehension, what did you think i meant by those keyword exist in rust.you can continue n and break n in rust...>I am smart, that's why I make $630k salary (tc 840) in hft.kek, imagine believing making a lot of money proves that you are smart, if anything now i think that you are a retard even more.>>109398103>is this you? no
>>109398105>don't go and pretend that not half of c project don't implement their own shitty version of a hashmap for their types because c lacks...>others suck therefore you donuh-uh>but what about metaprogrammingy-yeah, what about it?>you can do your sql queries, and at build time the compiler will query your databse to check for types mismatch between your quer...no, thanks. thats more code surface to maintain for the good compiler peopleid rather have them focus on their optimization magick>basic less so but i'm sure>i'm surethe irony is that i wrote a lil bit in basic myselfunderstood none of it cuz i was typing from a codebook into my ataribut i did, its kind of a badge of honor for me to have interacted with tech from such a span>i've used it in one of my first jobs and honestly i don't realy like it, i do get it has improved a lot since but still, not my cup of tea,100% legit>i've probably been an engineer longer than you have lolim self taught so i dont qualify as a proper engineer.doesnt prevent me from inventing maths or implementing shit that goes faster than whats considered the fastest by academiaim not saying im superior, i butt my head against holes in my knowledge all the timebut i know im good bc ive done good shit in programmingand you just cant argue with numbers>metaprogramming is basicaly writting code parse / write / edit code, but instead of a shitty script outside of your project, it's built into your language and can be part of libraries.cognitive noise. another interface you have to learn.
>>109398131>you can continue n and break n in rustNo, you can break to labels which is terrible and ugly. Numbers only.
>>109398158>nuh-uhyes, it does.besides you are sleeping on decades of research allowing better performance than naive hashmaps.and if you are telling me you are doing it for every projects again then that's all the more time and productivity lost.>y-yeah, what about it?you didn't even know what it was 10 seconds ago.>no, thanks. thats more code surface to maintain for the good compiler peoplemissing the forest for the tree again.also the compiler itself is less code because you can do more things in less sloc, so also less code to worry about.>id rather have them focus on their optimization magickthey also do, in fact because they are more productive since the language is better, they have more time for optimisation magick as well.>its kind of a badge of honor for meit is, but at the same time basic kind of suck.i played a lot with windows's batch when i was a kid too.>im self taught so i dont qualify as a proper engineer.i'm also self taught (started at about 9yo) but now i'm working as a senior.>doesnt prevent me from inventing maths or implementing shit that goes faster than whats considered the fastest by academiapretty cool but also a pretty low bar depending of what it is, academia is kinda retarded lol.>but i know im good bc ive done good shit in programmingsame, i don't doubt it, still, i think you'll probably appreciate rust if you take the time to learn it, or don't at that point i can't realy convince you, all i can tell you is that i'm not going back to c / c++, been like 6 years since i switched to rust and it only gets better.first few months were weird though ngl, but now it's just so much comfier to use than c/cpp ever was, it just doesn't get in my way, things that would be whole files in c are now one liners and i can still write some inline assembly whenever i feel the need to.
>>109398158>cognitive noise. another interface you have to learn.you don't have to learn it to use it, you can profit from libraries that do use it and would not be possible without.and even then, you can write some basic macros without having to learn much, a simple 10 line macro can save you hundreds of lines of boilerplate, you don't necessarily have to go deep.also it's not a lot to learn, unlike c where the preprocessor is a separate language, in rust proc_macros are just rust code.there are crate to make AST processing more seamless, but it's still just basic parsing of an AST.>>109398175>No, you can break to labels which is terrible and ugly. Numbers onlyi think it's better than the random goto.also if you realy want to you can still use crates that use metaprogramming to allow you some arbitrary breaks / continue into a well defined scope, but that's retarded imo.anyway, for every C feature rust doesn't have i can name 10 rust has c doesn't.and that one isn't particularly useful nor a good habbit.
>>109398237>b-but the hashmapsgeneralist solutions are largely inferior from tailor made onesfor the simple reason that its memory accesses that are the slowest, not math operationsyou can optimize the fuck out of your process, but if it needs more discrete accesses to memorythen its gonna be slowerits a truthism that a generalist solution will always need more of these than even a semi decent tailor made one>>no, thanks. thats more code surface to maintain for the good compiler people>missing the forest for the tree again.idek what you mean here. more features == more features>also the compiler itself is less code because you can do more things in less sloc,pushed to the extreme, a perfect function is just calling a main that contains all your programit wont give you any information thoughless slocs != easier to maintain, young padawan*more readable* slocs == easier to maintain.>senior>has to be taught about legibility of code...hopefully its a larp, bc oterwise thats a very bad testament to the state of the industryits kinda hard to believe that i didnt work one day in the capacity of programmer for someone else, yet i have better operational organization than a senior...otherwise, yeah, i dont think im gonna learn rustit offers mechanisms which i wont be using anywayseven c is too strict for me sometimes
>>109398330>generalist solutions are largely inferior from tailor made onesthat's false, that may be partialy true for C but that's not a problem when you have a language that has metaprogramming built in.>but if it needs more discrete accesses to memorythen its gonna be slowerabstraction doesn't mean you have to lose that access.>its a truthism that a generalist solution will always need more of these than even a semi decent tailor made oneno, also, most c project's hashmap is a pretty naive implementation that doesn't make use of the last 2 decades of algorithmic breakthrough and that's just against the std's hashmap.in fact if you now look at third party crates, tons of them implement tons of branches for your various hardware, making use of avx, simd, inline assembly etc, things you'd never reasonably do on on most projects, now you can benefit from those perf improvments freely.and even regarding memory access pattern.in C you can do array of struct or struct of array (which is better).in rust you can also do both, but what you can also do, is have a normal struct and have it turned into SOA by a macro, now you have the code simplicity of an AOS but it's compiled to an SOA.
>>109398330>idek what you mean here. more features == more featuresmore features means you can do more in less sloc, thus gaining time / productivity, thus being able to spend more on that time improving performance.you just have to look at existing projects, the vast majority of rust alternatives to C software (that aren't a lazy transpile / llm rewrite) completly mogs the performance of their c counterpart, case in point, ripgrep.>less slocs != easier to maintain, young padawanit isn't *necessarily* easier to maintain, but it often is if the design is good.>*more readable* slocs == easier to maintain.200 lines of if loops and garbage are never gonna be easier to read than one sloc that's just a map().filter().reduce().and even outside of loop, rust code is generally a LOT more maintanable than C in part because of the compiler check.ie if you add a new enum variant, it won't compile if it's not properly handled everywhere, with C you can make small edits that leads to error somewhere else in the codebase >>senior>>has to be taught about legibility of codeyou think you are teaching shit but you are just wrong, because again, your opinion stem from a lack of knowledge on the topic, stfu with that arogant tone.>yet i have better operational organization than a senior.you sure think you do.maybe that's alcohol speaking.>otherwise, yeah, i dont think im gonna learn rustyou do you, but don't start pretending you know what you are talking about when you literaly don't.>it offers mechanisms which i wont be using anywayscool for you i guess, i guess you don't know what you are missing on so you don't understand that you are missing on things you would use.>even c is too strict for me sometimesrust is less strict than c, it literaly gives you more control over what the end binary is gonna look like.
>>109398379>that's false,thats entirely true, if you have two execution paths one for each different case thats already more memory trafficadd to it the weight of the conditional, etcyou should work with some c, its gonna give you an intuition of these things>abstraction doesn't mean you have to lose that access.its not about as much about abstraction as of the richness of the objects youre working withtake an idealized example:compare an array backed linked list to a straight up arraynow consider the traversal of the data, even if the linked list is sequential relative to addressingyoure still gonna be slower in the linked list because you have to maintain more information, which translates to more memory and so more pages, cache lines etcwere at the level of applied information theory, reallyspecial cases are faster to process because information is implicit>no, also, most c project's hashmap is a pretty naive...>others suck therefore you do too, pt. 2...>in C you can do array of struct or struct of array (which is better).depends on the pattern of access...>in rust you can also do both, but what you can also do, is have a normal struct and have it turned into SOA by a macro, now you have the code simplicity of an AOS but it's compiled to an SOA.non-essential bling
>>109398417>thats entirely true, if you have two execution paths one for each different case thats already more memory trafficnot if it only happens once at startup or even better compile time.>you should work with some c, its gonna give you an intuition of these thingsi literaly told you i've used C for over a decade before moving to rust...and even C is pretty high level, it hides a lot of what's going on, i've done my fair bit of binary reversing and exploitation and yea C is not the end all be all.>its not about as much about abstraction as of the richness of the objects youre working withwords words words.still cannot give anything specific, like the last time i asked about your usecase.>muh array linked list argumentcompletly irrelevant, you can use either in rust, there is no limitation in rust preventing or forcing you to use one over the other...>special cases are faster to process because information is implicitthat information can be specified at compile time even if you use an abstraction / generic type, the only limitation here is you.>>others suck therefore you do too, pt. 2literaly every cniles say that they are special snowflake and then you look at their codebase or just their hashmap and it's fucking trash.stop lying, don't pretend all your code is perfect optimized bullshit, don't go pretend you never took a shortcut with a less than ideal solution because it wasn't your main focus.if you had a better language, all your use of a hashmap would be near ideal because someone spent literal months optimizing the shit out of it instead of your naive implementation that probably can't even top a single 1GBps in reading data out let alone more.>depends on the pattern of access...fair enough, in most case SOA will be a lot faster though, especialy if you have a lot of entities.
>>109398417>non-essential blingnothing is essential, you could use brainfuck for that matter.however, if you want to save time / have a better end product in the same amount of time and effort, then yes, it is essentialalso again MISSING THE FOREST FOR THE TREE.
>>109398510>>special cases are faster to process because information is implicit>that information can be specified at compile time even if you use an abstraction / generic type, the only limitation here is you.lol. that was low iqtypical crab.for your own sake i hope youre a larp
>>109398528>ad hominem because i have nothing more to say.>for your own sake i hope youre a larpsays the guy that argues out of ignorance, you don't know rust, i do.you think there are limitations in places there are not because you think in a C way with c limitations, which rust does not have.
>>109398554nonoyoure just fucking dumb, anon
>>109398528>>109398554and anyway, there aren't a billion ways to make the ideal array type.an array is continuous memory, your abstraction doesn't matter.but the algo arounds do.and if you do a reduce, instead of having your own shitty implementation, you can rely on built implementations for which weeks of optimizations have been put into.also you could use a crate like rayon, and in a single line of code process your array using multithreading.cniles often just not use multithreading because it's a pita to do in c.
>>109398565no u
>>109398571*inhales*nuh uhh (while exhaling for extra asshole points)in fact, i was so right a couple posts agothat i think ima touch myself to celebrate>special cases are faster to process because information is implicitwere touching (molesting even) the fundamentals of realitythe nature of information itself
>>109398596go to bed you are drunk.you didn't make a single argument.you failed to mention any specifics i asked.>muh special casesyou failed to mention one, an array of some random data type is not a special case retard.but maybe you are a special case indeed.no point in playing chess with a pigeon, he'll throw the pieces out of the board and proudly think he won.goodbye.
Safe C++ made Rust obsolete
some anti-rust /g/eets have been shitting this board for 2+ years, and they are literally still fizz-buzz level, while trying, and failing, to cover that fact with some izzat larp posting.let that sink in for a moment lol.some young non-/g/eet boy probably went from zero coding knowledge to doing some serious shit in that same time frame.
>>109398617lolleyou want a thought example?imagine two arrays of consecutive numbersone randomthe other numerically orderednow think how much information you need to describe eitherand guess where that information lives when this applies to a program>im playing chess with a pigeonno, the relation is that im teaching how reality works to a crab.quite literally.
>>109397925It's 11. Rust 1.0 came out in 2015. Prior to that it had gone through so many design revisions that nothing prior to 2015 can accurately be described as the Rust programming language.
>>109398702then it spent 9 years in development and is still shityour pick, im ok with either
>>109398672>now think how much information you need to describe eithersize of your type * size of the array + optionaly some metadata of your array container type (array size information etc) + extra bytes of alignment if needs be.the information needed to describe either has nothing to do with wether they are sorted or not, maybe if you talk about some shanon entropy compression bullshit but that's completly irrelevant to performance or the discussion.and nothing prevents you from writting an array that's just as optimized as c if you have some niche usecase where you want to trim metadata.but you seem to autisticaly obsess over things that don't matter, you know, amdahl's law.in fact C has some pretty crap stuff as well, ie null terminated cstrings.>guess where that information lives when this applies to a programheap or stack depending of how the array was declared.now rather, let's say you want to sort your shitty array, which do you think is gonna be the fastest, your naive quicksort implementation you did in a hurry because you wanted to focus on other things or a built-in or third party implmentation that literaly had months of optimizations to be the fastest with cpu specific instructions etc.and that's just for sort.there are hundreds of different operations you could do with arrays, reduce, matrix mul, rotate etc.i can just do .sort() and call a function that has had hundreds of hours of human optimisation behind it.you will either write a shitty naive implementation, or spend hours writting hundreds of sloc for something that's not even half as good as just ".sort()".and the worse is, the next person that has the same problem will do the exact same thing, where it could just have been a library if your language supported generics better.let alone metaprogramming reliant libraries.>im teaching how reality works to a crab.pigeon thinks he knows how reality, let alone just the sea, which is the crab's domain.
>>109398717>is still shit>can't explain whyyou literaly do not know the language, your opinion is worthless.
>>109398759its self evident>my opinion is worthlessyes, and uneeded. im just pointing out the obvious>>109398755holy wall of textyou get btfod by the basic properties of reality anontheres no coming back from thisif i were you id sepulchre to restore honor to my family
>>109389508If it’s good enough for Brave, it’s good enough for me
Polonius waiting room
>>109398772>its self evidenti say it's self evident that it is not, and i know the language.what are you gonna do.>>109398772>you get btfod by the basic properties of reality anon>supposedly an array in C is faster than an array in rust because magic (even if the exact same asm is being generated)okay nigger, keep believing your lies i won't stop you.>theres no coming back from thisindeed, you proved your own limitations to all of us here.>sepulchre>the drunk tries to take some religious moral high groundyou can't make this up lmao.
>>109398853>what are you gonna do.probably have some sleep safe in the knowledge that musings about information theory are, for now, beyond the reach of crabs,as expected
>>109398880>muh information theory>can't even explain how that's relevant to the discussion.>yea trust me bro, C is faster be... becau... because... le ENTROPY or something.. totally not an argument pulled out of my ass without any relevancy to the discussion.retard.if the instructions are the same, and the data is layed out the same way, it'll perform the same, that has nothing to do with le information theory, i understand you like to use big words but you are only making yourself look like a (drunk) retard.you seem to think the cpu is made out of magic balls or something.
>>109398905that sounded almost sentient.youre making progress, anon!
>>109398928and you sound drunk, retarded or both.you haven't made a single compeling argument in the whole thread yet somehow you proudly think you somehow won some debate and you don't even see that you've just been farting in the wind.i truely pity you.
>>109389508It is tedious to write Rust code and AI doesn't work well with Rust, the advantages of writing a new project in Rust are usually huge but rewriting an old project in Rust is a waste of time and resources especially if the project works wellAlso Rust is not good for quick development and prototyping, if anything developing with Rust is a slow process, you need to write some good Rust code to get the promised performance
>>109389508I'm done with all these fucking meme languages.Rust, Zig, odin, hare, faggots retarded language of the week. No thanks.I'm sticking with Go. I have understood it finally that Go is the past/present/future.If it's good enough for Google, it's good enough for me.
>>109399002>It is tedious to write Rust codefalse, rust is really comfy to use imo.provided that you master it of course, all the ones finding it tedious are those that don't know it well enough.>AI doesn't work well with Rusti hate vibeslopping, but it's literaly one of the language it works the best with because the compiler will tell it why shit is not working.maybe try not using a 1B model.>but rewriting an old project in Rust is a waste of time and resources especially if the project works well.i agree.the only thing that justify a rewrite is that you want to add new features and it'd be more work to add them than to rewrite and add them.>Also Rust is not good for quick development and prototypingdisagree again, if you master the language it's pretty great imo.the only issue is that most people give up on it before getting to that point because the learning curve is pretty steep.if you master it it's an amazing language to prototype with, and it's also really good for refactoring due to the type system.>if anything developing with Rust is a slow processthe n1 reason i switched to rust (before llms were even a thing) was that i was a LOT more productive with it than c++, it just is a more productive language because it has more features, better tooling, better ecosystem, build system, package manager etc...>you need to write some good Rust code to get the promised performancethat's not specific to rust.also amdahl's law, you realy should focus on the hot paths.
>>109399241>If it's good enough for Google, it's good enough for me.google literaly made go to be able to hire jeets, it was literaly designed for retards.i rather write c than fucking go, using that language is an insult on anyone that's not retarded, it just keeps getting in the way everytime you want to do something.it'll refuse to build for a fucking unused variable or import (because you are prototyping something) yet will be perfectly happy you not handling errors.
>>109399271Go is fun.
>>109399295>Go is fun.well i'm glad you like it anon, but i really hate it.still, i don't want it to disappear either, i'll keep working with Rust because i find it very comfy, but in the end i guess people are different and they'll like different languages.trying to force gopher or cniles to switch to rust (or vice versa) is not gonna be a good idea.
>>109399321Hey, I like Rust. Idk if other Go fans do.
>>109399407i know one go fan that likes rust.he didn't like it at first then he learnt it and kind of made it his main but i think he still likes go.i never really liked go, i had a short term contract where i worked a bunch with it and yea i'm now declining anything that involves it.
>>109398652rust is literally a jeet language. it is pushed by corporations so they can hire less skilled programmers, i.e. jeets, for less consequences.
>>109396188Run cargo tree on any end-user repo. Every dependency with a (*) is a de-duplicated repeat
>>109394870>a lives longer than b>aka as long as b exists, a has to exist too>way more complicated
>wake up>over 200 repliesthe troon discord was notified, huh?
>>109396188>This requires someone actually understanding the dependencies in order to make them overlapNo it doesn't. Packages follow semver, the default version resolution mode is "first non-zero must match plus semver compatibility for non-zeros after", which is semver that's slightly more permissible only in the under-1.0.0 range. So if something depends on "2.3.4" it actually is fine with the range [2.3.4, 2.0.0). If something else in that same graph needs "2.5" of that same thing it will just get deduplicated to one instance of with the version 2.5.x.Cargo is a source code distribution system, not binary package one, so all of the version resolution happens locally for the one building (or is taken from cargo.lock cache of the root project to not re do it on every build), none of the devs of your dependencies need to think hard about overlaping with others. Any random project will have A TON of deduplicated dependencies. Brb imma go and check
>>109397079What the fuck are you talking about retard, there have been no breaking changes since 1.0 like a decade ago
>>109397425Filtering out retards that can't learn new things.
>>109400213No, it's the fucking anti-rust retarded cnile, as always
>>109400246>>109396188>cargo tree | wc -l1329>cargo tree | grep '*' | wc -l821de-duplicated deps have a "(*)" on their linekeep in mind this is UNDER counting it, because cargo tree only prints the dependencies of libshitto x,y,z only once, every next occurence will just feature libshitto x.y.z (*) without the entire tree under it. In other words deduplicated packages only show up if they're a 1st order dependency of a non-deduplicated package or the 1st occurence of the deduped package>cargo tree --no-dedupe | wc -l26221619>inb4 lmao 26 mil!I know nobody on /g/ has formal education, so I'll explain it to you, this is what happens if you unroll a DAG with a lot of splits and merges into a tree, combinatorial explosion
>>109400246You had to go on this bs tangent just to make an argument, do you have any idea how ridiculous you are? Nobody fucking asked for an explanation of how cargo works. I can't believe you actually think overlap means not having duplicated copies of the same exact code. Thats not even close.Also if you are importing 20 or so different micro libraries with preconfigured interactions with other micro libraries it might as well be a macro library at that point.When you actually use micro libraries themselves, you just route information over to them and let them do their own thing, that's not overlap. The micro libraries need to connect to other libraries directly for it to have overlap.And you have to either edit every library for that, or use a macro library.
>>109400285As someone who has written in assembly. Assembly is simpler than rust.What is even the point of a programming language that is less functional and more confusing than what it is abstracting?
>>109396595no, that is one of the best parts of rusts - hygenic macro system. it's clean like latex clad ass of bdsm chick sprayed with Vivishine whereas C macros is pajeet smeared with feces. choice is obvious
>>109400557>Assembly is simpler than rust.Assembly is simpler than every high level language.>less functionalRust is a more functional language than assembly.
>>109397856Maybe he means breaking with value? That's also supported in Rust.let result = loop { if b > 10 { break b; } let c = a + b; a = b; b = c;};assert_eq!(result, 13);
let result = loop { if b > 10 { break b; } let c = a + b; a = b; b = c;};assert_eq!(result, 13);
Rust anons, praise more hygienic macros! The clean design! Omg! No context collisions! Metavariables! Erection! Hygiene! Shiny! Clean! Sex! Hnnnng... macro_rules! Ejaculation!
>>109400285>camps the thread for 5 hourskekimagine being so mad for so long
>>109401014what is this? a pic for ants?
>>109393258Standard deviation of IQ is usually around 24 so if youre talking about an equivalent scale, sub 50 IQ would be like 148,800,000 people worldwide
>>109398755there was a performance issue in QEMUs 9pfs driver that was literally what this anon is saying. cniles just reinvent the wheel (poorly) every time they encounter a basic bitch problem or go full GNOME and use glib like a sane cunt anyway.
>>109398175>No, you can break to labelswhich is the same thing, but better.
>>109400729no, he's retarded and literally wants to doloop { .. loop { // break out here loop { break 2; } } ..}because he's too retarded to use labels.
loop { .. loop { // break out here loop { break 2; } } ..}
