Comments by "Vitaly L" (@vitalyl1327) on "ThePrimeTime" channel.

  1. ​ @litpath3633 degree is a way to ensure that the graduates at the very least did not skip the basics. Because "self-educated" always do, their knowledge is full of holes.
    3
  2. 99.5% of my code is generated by a compiler. So what?
    3
  3.  @kodingamedev  yes, Nim is a worthy language, macros done the right way.
    3
  4.  @camdustin9164  you don't understand macros then. A language with macros can be turned into ANY language you can imagine. Any feature that exist and even those not yet invented can be added, or a language can be completely turned into something else. Therefore, any language with macros is a superset of all existing and not yet existing languages. Any language without macros is nothing in comparison. Now, if you managed to live your entire life without a need for macros, you're programming wrong. Badly wrong. You avoided the most powerful tool for building abstraction. And, no, language popularity is not a measure of language quality in any way, because vast majority of programmers are incompetent monkeys.
    3
  5.  @j.trades9691  what performance losses? DSLs are zero cost abstractions, unlile pretty mucb everything else. And, unlike everything else, DSLs allow to eliminate complexoty. Not just reduce, but elimiminate. If your problem can be expressed in terms of compilation (and DSL methodology is exacrly what allows to discover if it can be done), it is a solved problem. Compilation can be split into a sequence of trivial steps, as trivial as you want, thus eliminating complexity altogether. As for learning curve - see above on complexity.
    3
  6.  @maelstrom57  name a single good thing that appeared in the standard since 1.6. Everything that followed made it just more complex and less suitable for static analysis.
    3
  7. How is it confusing? Your theoretical GFLOPs are pointless if you cannot feed the new data fast enough. And memory bandwidth available shows you that you indeed cannot - hence this 5.25 GFLOPs of data in + data out. And even this is probably not sustainable. Would have been hilarious to watch Primeagen discover all the scrambling, space-filling curves, local bank-wise distribution, etc. that are used in image processing algorithms to improve cache locality.
    3
  8. @MariusSchwendtmayer  I am the one who is using the trash web "developers" produce, so I have all the moral rights to mock them as much as I wish, as I am on the receiving end of their institutional incompetence.
    3
  9. Just use proper static code analysis tools to map the flows. No need to watch them step by step in a debugger.
    3
  10.  @RedstoneSea5124  yes, any LSP based mode can do it.
    3
  11. Is not it a point of this whole challenge? To optimise the memory access pattern (it need to write, not just read, and the output may not fit L3)
    3
  12. LLMs can solve problems far better than the vast majority of humans already.
    3
  13. ​ @septuleptum garbage in, garbage out. The entire web foundation, all the w3c standards, including JS itself, are utter garbage, so no matter how hard one try to build something pure or to create a meaningful framework, they all end up producing pile of crap. Nothing will ever save web from overengineering. And the quality of the web "developer" hordes does not help.
    3
  14. Just stop this backend/frontend nonsense, get back to local UI applications. Web is a massively overengineered pile of dung, and there is no saving it.
    3
  15.  @DrQuantumInfinity  Huh? Quantum physics is a textbook example of how the right language and the right abstractions simplify an insanely complex thing into something uttterly trivial. Bra and ket rules.
    3
  16. ah, how cute, web coders think that anything that is not web is a "niche". Look, if your precious web suddenly disappears, the world will just become a little better, and that's all. If, say, MCUs disappear, civilisation will collapse.
    3
  17. well, developers who build on top of vast third-party "frameworks" are just doing it wrong. In the vast majority of cases using a third party dependency is rather a liability than help, and can hardly be justified. The programming culture went down the gutter in the recent decades, with developers thinking it's acceptable to drag dozens or even hundreds of third party dependencies for even tiniest of functionality. It's all utterly disgusting. So, yes, attitude towards debuggers is a good marker of what culture programmer is more aligned with, and yes, I'm using this as an interview question to filter out undesirables.
    3
  18.  @brinckau  people like Bellard are so far on the tails of distribution that it does not even matter. You cannot possibly hope to ever see one. Also, this is not a normal distribution.
    3
  19.  @monad_tcp  of course, the whole point of MISRA C is that the entire code base must conform. If you have a tainted part, it breaks safety guarantees for the rest. In automotive and aerospace it is very often a requirement, no tainted code is allowed, no matter what.
    3
  20. found a pathetic flaturd
    3
  21.  @leisiyox  always thought that Prime is, like, the embodiment of the average developer - he's doing web and all that. That's why I find his streams interesting, it's like the voice of the regular people.
    3
  22. Cope more! Reality will bite.
    3
  23. Why cannot you push boundaries and explore weird stuff in your day job?
    3
  24.  @sumansaha295  sorry, mate, I'm afraid it's you who have no idea what you're talking about. C and Java may seem different only to those who don't know anything else. Both are eager-order imperative languages, both relatively low level. Rust, on the other hand, was drawing quite heavily on the ML family, which is, in turn, worlds apart from the likes of C and Java.
    3
  25.  @RotatingBuffalo  as for macros in C, no, they're not actual macros, and they're not really part of the language, preprocessor in C was always very separate thing. Compare that to procedural macros in Rust, or to C++ templates.
    3
  26.  @Aethenthebored  they are similar in all possible ways, and this is exactly why code monkeys have no problems moving between those languages, as tbey are all one language in different dressings. And when they are confronted with an actually different language, aa with the post we discuss here, they have hilarious meltdowns.
    3
  27.  @HrHaakon  lol, then Excel "macros" are macros too. But still they are not the macros in a PLT sense, just like C "macros" are not called macros in PLT. Anyway, in a practical sense, anyone with an experience with C "macros" have no idea what real macros are and how to use them, and therefore is going to be baffled by Rust procedural macros, or by Common Lisp macros, or by R5RS hygienic macros, or by C++ template system.
    3
  28. ​ @theheadpriest look, I understand that PLT is hard and that code monkeys do not even study it in the first place, but yes, imperative eager languages with separate statements are pretty much all the same. You're confused because you don't know of any other types of languages to begin with. When you look from a perspective of Agda or even Prolog, all the statement-based languages are indeed the same.
    3
  29.  @Legac3e  yep, and this is exactly what is not happening in the industry. And if I am to chose between regulating the crap out of everyone and not eegulating at all, since selective regulation can be very hard to apply, I am for the former.
    3
  30.  @Legac3e  exactly, if we have to pay by slowing down the pace of churning out software (most of which was not necessary in the first place) for making less likely the incidents like Toyota breaking, Postoffice sub-master scandal, 2019 NHS trusts cryptolocker scandal, the recent air traffic control shambles, etc. - it's a tiny price to pay. Also, you never know where the software will end up. The coders who wrote the Postoffice system were mere code monkeys crapping out very regular CRUD. They screwed up even there, and it resulted in many real people deaths and broken lives. You do not have to work on a high risk control software to be able to cause massive damage, a mere CRUD can kill.
    3
  31.  @SimonBuchanNz  not really, all variables were global all the way until Algol60 introduced recursion, then local variables became a separate storage class, allocated on a stack, or heap frame, or whatever method you use to implement recursive calls.
    3
  32. ​ @Salantor  Travis is a perfect example. 99% of what he's saying is an utter nonsense.
    3
  33. Youtube auto-mods delete comments nearly randomly.
    3
  34. Why do people try to drag Go into a C++/C/Rust discussion? Let Go compete with the other toy languages like Java, C#, Python and JS. It cannot be compared to C++, for it does not fit the niche where C++ is the most useful.
    3
  35. Huh? The very nature of problem solving is in finding ways to make it easier.
    3
  36. And what's wrong with having a C compiler to parse your headers? Parsing alone is not enough for any language, and compiler is exactly the tool that allows to introspect a code and do whatever you want with it. It's just plain wrong to ever use any other tool than the language compiler to parse its code. Also, writing a compliant subset of C compiler (only the parsing and typing frontend) is a relatively trivial task anyway.
    2
  37.  @TheRealXartaX  so, you have no idea what PEG or GLR is, yet you're soooo confident.
    2
  38.  @bigdogdman1  again, there is never a reason to use regular expressions, because there are always much better tools, for every use case.
    2
  39. Same thing for me - Emacs enrages me pretty much on a daily basis, but I could not find anything better than Emacs.
    2
  40.  @VuxGameplays  wtf are you talking about? I am in the field for more than thirty years, and never once touched the web, not a single tiny thing. Most of the people I work with also never did anything related to web. Why new developers believe that there is nothing but web?
    2
  41. It's a high time all kinds of web "developers" are automated out. They were never real developers anyway.
    2
  42.  @lukivan8  not in a slightest. If someone "can" do surgeons work, but don't have a licence, don't have obligations to a regulating body, don't have all the necessary personal responsibility safeguards, he's not a surgeon. He's an educated dilettante. Same goes for the engineers. It does not matter how hard you self-studied, how high your IQ is, but nobody will ever let you design a bridge if you're not a real engineer, with obligations to regulating bodies, with personal responsibility, etc. Now, same goes for software engineers too. I am not going to believe you if you'll claim you, personally, never were a victim of rogue software engineers. That you never suffered from using bad or outright dysfunctional software. You did. Everyone did. And nobody should have suffered, if only this field was properly safeguarded, like medicine and other engineering fields. It's a high time we kick out all the "self-taught" who cannot pass certification, that we introduce governing bodies, that we introduce personal responsibility for any consequences of the bad decisions made by software engineers.
    2
  43.  @lukivan8  Diploma does not matter much, true, what matters is professional regulation organisations and enforcing of personal responsibility. This is what differentiate classical engineering from the mess that the modern software engineering is.
    2
  44.  @suede__  another one with reading comprehension issues? Firstly, yes, any competent engineer must have a broad expertise in an array of languages. And, secondly, any decently sized project is way, way bigger than an effort of learning some language or framework. And, yes, a lot of really awful people are in this industry and I'd do anything to push them out. If the trash coders who wrote Horizon were gatekeeped away, people would not have suffered. Same goes for every single software failure that did cost a lot. This industry need professional standards similar to medical and civil engineering fields.
    2
  45.  @svuvich  primeagen is far from an example of a.competent engineer. And this industry absolutely needs a lot more gatekeeping. Ideally, civil engineering level of gatekeeping.
    2
  46.  @svuvich  that's a very good analogy. Just an hour ago I saw a guy being helped by paramedics after he fell down the stairs on the Tube. A tile on the stair edge was wobbly and he lost his balance, tumbled all the way down. Coders who wrote Horizon were also not much more than tilers, yet they caused immesuarable harm to hundreds of people. Quality standards are important regardless of how small the job is.
    2
  47. Yes. Pretty much everything in the modern computing is overengineered beyond any measure. And not a small part of this is thanks to the self-taught and otherwise insufficiently trained engineers who know no better.
    2
  48. You mean, a way to add 2 and 2 different from Peano? Right. Sometimes performance is more important than theoretical purity, that's why we have so many different ways to add 2 and 2.
    2
  49.  @gregorymorse8423  No, it is not 1 clock latency, there is not a single CPU that can do it. All the CPUs do 2 or 3 iterations, each made of a floating point multiply-accumulate.
    2
  50.  @123456crapface  you're funny. Real software deals with a real data, and you won't trust your real data to a cloud.
    2