https://www.youtube.com/watch?v=U46fJ2bJ-co
deprecated by Rust
>>108869053Rust is ugly as fuck
>>108869103Not just ugly, it's uglier than C++ and JavaScript combined
>>108869053Rust is deprecated by AI.
Programming languages are suppose to be an abstraction away from machine code not more complected than it.
>>108869126unironically, kekai handles c++ just fine and there's mountains of training data with an already healthy existing ecosystemno catching up to that
What level of hell do you think it will burn in?
>>108869053fpbp.
>>108869053what is rust :S
>>108869904I don't fucking get this guy. “A cause for concern (not panic)” as if everyone thinking the language is shit and horrible is just a minor inconvenience. How does he wake up and take himself seriously while pretending this crap has any value?
>>108869904The keyword is languages.Rust is a pedophile cult.
>>108870218Interesting. What about Pascal and Ada?
>>108869995How dare you, sir?How dare you?You are a monster.
>>108869034Official C++ events are full of trannies and feminists. The C++ community is not a serious community and they are not professionals.
>>108869034The first language. Everything from C with a few extra tools. Ignore the OOP cancer and use it functionally and it's perfect.
>>108869034What the fuck is that haircut
It's called the come on guys it's not so bad.
Why does everyone call C++ bad when most games are being developed in it?
>>108869034we should work on memory-safe evolution of c++
>>108871693they really don't make em like they used doold school no nonsense white dude programmers with 3 kidsthey were the apex of the programmer specieslook what we have now: trannies and indians and chinese
>>108870859Doesn't matter, all that matter is that he chose this haircut on purpose, which tells us all there is to know about C++.
>>108869126Sorry best AI can do is python. It can't into low level coding at all, it can't even do C.
>>108869132>mountains of training dataMountains of buggy trash using conflicting dialects (fno-exceptions) and subsets and even custom replacements of basic functionality such as rtti.Idiomatic C++ does not exist.
>>108871693Just adopt the unix philosophy, no memory just text streams.
>>108869034OP is this your channel? No clue who the interviewer is.>>108869053C++ is deprecated by C. The only valid reason to pick C++ over C is if you don't want to write your own data structures, and most C container libraries are bad. C++'s STL is pretty bad too but at least it's reliable and very well documented.But dynamic arrays + hash tables are 90% of what you need anyway.Rust is a terrible language in every way possible. Crates are like npm hell all over again except worse. It's just degenerate. Zero stability, can't dynamically link in practice, depends entirely on one single toolchain (LLVM), extremely brittle low quality dependencies being pulled from other crates, and so on.
>>108869053Just like your penis. Google tranny.
>>108869034>admits mistakesis he dying soon or what?
>>108869034Sepples set back computing by 30 years.
>>108871822you are a nocoder and a retard
>>108869904Haha, they still leak, but you have no idea, hahahahahhahah
>>108872775leaking is memory safe, so rust doesn't prevent it
>>108870817Just use C.
>>108871705True.
>>108873074C++ is C but better. There's no reason to use C.
>>108869034>I'm so sorry that the porsche I donated to you has a dent on it, truly, I have failed...
>>108869034>2 days ago
>>108873140Designated initializers.
>use our convoluted C++ zero cost abstractions that can be compiled away during the hours long compile, instead of using light C abstractions that compile fast to begin with
>>108869034He actively killed it along with the committee that hamstrung the development
>>108873140>better initialization that looks like json except it's even better>every C library is 10x times easier to use and read because C APIs document themselves>code is easier to read>much faster compile times>no move semantics nonsense, a symptom of the disease that is traditional OOP where everything must be a class>no template bloat that is impossible to debug, macros are enough and unironically easier to read>every object is either an id or a POD struct just like god intended>better and much simpler memory managementI could go on, but the only good parts about C++ are all the parts it inherited from C. The further it strayed from C, the worse it got. People use C++ or now Rust because they never tried C. It's sad.
>>108873705C++ 20 adds this I think.>>every C library is 10x times easier to use and read because C APIs document themselvesThere are 1000 times more C++ libraries. Sorry but C++ won.>>code is easier to readIt's the same fucking code retard. C++ is C.>>much faster compile timesDon't care.>>no move semantics nonsense, a symptom of the disease that is traditional OOP where everything must be a classMove isn't OOP at all, it's functional programming where you try to refrain from mutation. C++ doesn't tell you how to code. If you use OOP it's your fault not the language.>>no template bloat that is impossible to debug, macros are enough and unironically easier to readIt must be opposite day. Macros should never be used and templates are great.>>every object is either an id or a POD struct just like god intendedAgain, if you want everything to be a struct then go for it. C++ is C.>>better and much simpler memory managementSimply wrong. C++ added better pointers.
>>108874058Name mangling, no de facto standard ABI and fuckhuge compile times are barely scratching the surface of the problems that Sepples has, but C does not have.
>>108873520C++ has had that for yearsC++ also has templates and lamdas which opens up all kinds of compile time optimization opportunities
For me it's Gnu Make, C and tcl/tk.It's pretty sad that there has not been a better GUI library.POSIX could have defined API for GUI, they really should have defined more optional APIs for a more civilized age.
>>108874697These are legitimate complaints and I like C++. Compile times have never personally been a problem but I'm not working on really big projects.
>>108874697>>108874780Compile times shouldn't be a problem even on big projects. The only reason you'd get long compile times is if you have no idea how compiling works and you fucked up your namespaces. In C languages compile times are mostly a project structure problem, it's easy to screw up and then changing one function makes your entire project recompile. The solution is to not be a jeet and read a book on good namespace structuring.
>>108874814There is no technical reason why adding a private function to a class should trigger recompilation of every translation unit that uses the class.
>>108870608Pascal is based and deserves more love. Ada is a retarded abomination but still better than rust.
>>108871542People use C++ because nothing else has the same cost/benefit tradeoff, even though if even a fraction of the money and time went to other languages as had been invested in C++, that would never have been an issue.Namely, C++ allows tight control over execution and particularly memory patterns, which is important because games are soft realtime systems with high performance demands. Other languages like C and asm also fit this niche, but asm makes cross-platform development difficult, while C makes managing large projects with many contributors painful. The problem with rust is that games require a very fast iteration loop to playtest mechanics and figure out quickly what works and doesn't work. Rust does not allow for this due to the way the type system works (see for example https://loglog.games/blog/leaving-rust-gamedev/). Furthermore, the C++ ecosystem is far more developed than anything else, lowering greatly the time, cost and effort to implement a new feature compared to some other options.
>>108874903Cool. Now, COPE-splain how bjarne pretending not to see those three languages in the list was a total coincidence.
>>108875276No cope, bjarne is a retard but he's also an easy read. Everyone knows why he cherry picked the languages he selected.
>>108874903https://www.lysator.liu.se/c/bwk-on-pascal.html
>>108871542Just because something is popular does not make it good.I would expect every person past elementary school age to understand this concept.
>>108873705>better initialization that looks like json except it's even betterDesignated initializers in C99 and later are truly an amazing feature, no surprise every C project that has been maintained during the past 20 years eventually adopted them.Unlike most Sepples shit, designated initializers are a true zero-cost abstraction: you can initialize fields out of order without having to worry about the underlying layout, and this provably costs NOTHING at runtime. The cost only has to be paid "once" in the compiler to actually implement the feature.>no move semantics nonsense, a symptom of the disease that is traditional OOP where everything must be a classIt's not a consequence of OOP per se (almost no other language has this nonsense), it it was simply the realization that shallow copies (which is what move semantics actually are in practice), are sometimes useful, and that "hardwiring" the assignment operator to always perform a deep copy is not just wasteful, but also confusing.To further illustrate this point, there is no distinction between copying and moving POD types: it's all just a field-by-field copy of the object all the same.>no template bloat that is impossible to debug, macros are enough and unironically easier to readAgreed, I stand by my unpopular opinion that macros are at least as good as templates for 99% of the shit they are actually used in practice (type generic containers and functions), and can even do more (they are not limited to being function/class skeletons). Are they as sophisticated, as, say, Lisp macros? No. But they are useful nonetheless.
>>108869904Based Bjarne exluding the trannylang from the list.
>>108875552Yes. He is /our/ nocoder.
>>108876000>a ship analogy
>>108875484Nearly all the arguments are pedantic at best, and those that are legitimate were not actual at the time of these writing but were already about 11 years out of date. Might as well compare proto-C (which is untyped) and then suddenly half the arguments get reversed. It is particularly striking that he calls pascal not serious compared to C because at the time it was still actual that C was not seen as serious compared to asm.
>>108876076getting into details is not really necessary. any real language is better than the pile of endless shit that is C++.C, (free)pascal, Ada, Rust, Zig (wip), or whatever, all of those would be preferable over C++, when you have the ability to choose (which is not always the case obviously).and that's assuming "managed" languages are not acceptable for the use-case. if they are, then you can add a bunch of other languages which would be preferable over the gargantuan 4 decade old meme.
>>108875539macros are pure evil, a hack at best, and templates are only good in theory, the implementation is horribleI've been programming in .NET/C# for over 2 decades now and while the language is very comfy (they keep adding bloat, tho), I long for the simplicity of C
>>108875539>designated initializers are a true zero-cost abstraction: you can initialize fields out of order without having to worry about the underlying layoutC++ got designated initializers too but they left out the most important feature: out of order initialization. If you change member order in the definition it breaks all your initializers.
>>108869904imagine doing what NSA tells you lol
>>108870608FreePascal is fucking based but without some big marketing cash investment normie """developers""" won't ever consider it.
>>108874924I'd just like to interject for a moment, what you call C++ is in fact no exceptions C++ or as I've recently taken to calling it, not C++.
>>108877068imagine fake half-quoting the NSA like a retard because you COPE with your insecurity like a literal child lol
>>108869034I don't watch videos which such clickbaity thumbnails, sorry.
>>108874924>C++ ecosystem is far more developed than anything else, lowering greatly the time, cost and effortREEEEEE
>>108874847Sounds like a you problem for using OOP and not managing your translation units. Also there are heaps of reasons for wanting to recompile everything for instance cross unit optimisation. Get gud scrub.
>>108874924tl;dr C++ is perfected after decades and you cannot compete and will never be a woman
stuped+evil
>>108876921And how is any of this relevant to the post or the article it replies to, retard?
>>108878308No, C++ got fucked up after decades. It's like python: the most horrible language known to man, with the worst 3rd party ecosystem ever seen, but still would take a couple of weeks to rewrite the essential parts of 2 of the most important libraries in the ecosystem into a language that doesn't suck so you can use it elsewhere, so nobody is willing to do that.
>>108878652Why is 3 worse than 2?
>>108878291Just admit it's a problem.One of the perks of not using Sepples is also not having to work with people like (You).
>>108877019I can't conceive why the fuck they did it. It makes no sense at all.
>>108876988>macros are pure evilJust because you read this many times on StackOverflow doesn't make it so.
>>108878701I don't see the relation since I never said anything about that, but since you asked.Every subpoint version of 3 is both forward and backward incompatible, so you can't just use a venv, you have to use an entire local install of python and all your package dependencies and communicate between python modules using the shell (i.e. os.system or subprocess.check_output etc.) since they won't run with the same python version and an import.Moreover, 3 is slower than 2 by about 25%, which is quite the achievement given how slow 2 was.
>>108878708>Just admit it's a problem.Jeets are the only problem. noob
>>108878652Python and C++ are great together. You just hate anything that's popular noob contrarian.
>Jeets are the only problemYes, but not in the way you think.https://yosefk.com/c++fqa/fqa.html
>>108878881They're not even compatible together and you failed to even provide an argument, not to mention the non-sequitur nature of your bullshit post.
>>108869132>ai handles c++ just fineNot defending the rust troon but LMAO no it doesn't.
>>108878938True, but to be fair humans can't handle C++ either.
>>108878907>They're not even compatible together???Go back to india.
>>108877260ok zogbot. still not using your pozzed faglang.
>>108879064>ISO is not zog
>>108878647it brings the subject back to this thread. it's the blogtard reply that is irrelevant.Pascal doesn't have to be good. Being less bad than C++ is sufficient as a starting point.
c++ and rust are both gay and retarded, hth
>>108879176>brian kernighan>blogtard
so he was serious with all the C++ bullshit?lol retartd
>>108879859That's the saddest part. Sepples wasn't a joke after all, as the satirical cat-v interview suggested.
>>108879436>CS celeb means not tardhow is the weather in jeetland today?it's on the colder side around the Mediterranean.
>>108878991he's right, retardthe existence of pybind11 does not mean the two languages are compatible with each in a semantic senseotherwise literally every language is compatible with each other thanks to bindings, even fucking COBOL with javascript
>>108876000Based, even more based
>>108880051Speak Christian.
>>108874745c++ designated initializers suck, you have to put the fields in the same order they're declared in the type so any reordering of the fields later causes warnings everywhere else
https://www.youtube.com/watch?v=7fGB-hjc2GcC++ is an abomination. The only reason to use it is if you have to for work.
>>108880112>every language is compatible with each other thanks to bindings, even fucking COBOL with javascriptYes, and? Also go back to india.https://docs.python.org/3/extending/extending.html
>>108880400>retardo posts self-own>blames indiaEven poos are better at this than you.
>>108880588I accept your concession.
>>108869904He’s in denial
>>108880400so everything is "great together", moron, what's your point?if you actually work in this industry, then expect an indian to replace (You)
>>108877260Mars rover died because someone left an unwrap() in there and it panicked.
>>108881114>so everything is "great together"Did I say everything? Pretty sure I said python and C++. Maybe learn to read English and then go back and read it again, or just stop posting on an English speaking website.
>>108881236>Pretty sure I said python and C++.did I say you didn't say that? pretty sure you lack the mental capabilities to grasp where your argument logically leads if you believe python and C++ are great together. maybe programming isn't for you after all
>>108871776very insightful, sir
>>108881286Your strawman has nothing to do with logic. You've got the logic of dust mite.
>>108881750I accept your concession
>>108869034C++ is double plus goodC# is 4 plusses stacked on each other, so double double plus good
>>108870859backward compatibility
>>108871822why are crates worse than npm
>>108869034Europeans should be banned from designing programming languages.
>>108882201The best designed programming languages are all european. The worst are all amerilardian.
>>108882201You should be banned from designing posts on the internet.
>>108881138stat tuned. you will be presented in the "nocoder" section of my next article "On Anti-Rustists And Their Cries". it will actually present 3 types of coders, of which you are not one. i'm keeping the title of their section a surprise.
