C++ is the future!

C++ is the future!
https://www.phoronix.com/news/CPP-Linux-Kernel-2024-Discuss
>"Now, "why not Rust"? First of all, Rust uses a different (often, in my opinion, gratuitously so) syntax, and not only would all the kernel developers need to become intimately familiar to the level of getting the same kind of "feel" as we have for C, but converting C code to Rust isn't something that can be done piecemeal, whereas with some cleanups the existing C code can be compiled as C++.
>However, I find that I disagree with some of David's conclusions; in fact I believe David is unnecessarily *pessimistic* at least given modern C++.
>Note that no one in their sane mind would expect to use all the features of C++. Just like we have "kernel C" (currently a subset of C11 with a relatively large set of allowed compiler-specific extensions) we would have "kernel C++", which I would suggest to be a strictly defined subset of C++20 combined with a similar set of compiler extensions.) I realize C++20 compiler support is still very new for obvious reasons, so at least some of this is forward looking."
@
rusttrons BTFO!

Beware Cat Shirt $21.68

Rise, Grind, Banana Find Shirt $21.68

Beware Cat Shirt $21.68

  1. 1 month ago
    Anonymous

    top class c++ dev just wasted his time writing useless code

  2. 1 month ago
    Anonymous

    i would love to see this because of the endless Cnile seethe it would cause

    • 1 month ago
      Anonymous

      what did the rust troony mean by this?

      • 1 month ago
        Anonymous

        just look at the homosexual below you and above me to see what i mean, lmao

    • 1 month ago
      Anonymous

      based, making the Linux kernel a C++ project would also improve it's velocity with respect to changes, because some abstractions are actually fricking useful. (classes are abstractions)

      and C projects over a certain size can benefit immensely from C++

  3. 1 month ago
    Anonymous

    C++ is cancer. I have never seen a C++ project that was not an incomprehensible mess. For a prime example see the llvm/clang source code.

    • 1 month ago
      Anonymous

      >I have never seen a C++ project that was not an incomprehensible mess.
      SerenityOS is pretty readable, best Ive seen

    • 1 month ago
      Anonymous

      >I have never seen a C++ project that was not an incomprehensible mess.
      Dear ImGui

    • 1 month ago
      Anonymous

      It's not my problem that you can't into classes and you don't understand the difference between templates and TMP.

    • 1 month ago
      Anonymous

      what the heeeeeeeeeeeeeeeeellll

  4. 1 month ago
    Anonymous

    >rusttrons BTFO!
    How are they BTFO? Rust already has the greenlight. C++troons are just trying desperately to get attention again as their language is clearly dying.

    • 1 month ago
      Anonymous

      The article which was linked, and the recent comments from Torvalds indicates that he doesn't expect rust to go much further than driver support. C++ probably won't get into the kernel. But you've been able to write C++ drivers for Linux for a very long time.

      • 1 month ago
        Anonymous

        There is no fno-exception+fno-rtti dialect driver in mainline though.

        I think it's likely a Rust driver gets mainlined before C++.

        • 1 month ago
          Anonymous

          Ok and? It's still a driver.

      • 1 month ago
        Anonymous

        Nice cope buddy. Whatever helps you sleep at night.

        • 1 month ago
          Anonymous

          He pretty clearly lays it out very clearly in his recent interview. Basically rust hasn't gone anywhere since it was introduced. There aren't any serious plans to go beyond a few drivers. Maybe mainlining one or two drivers which already have c implementations. This astro turfed rust shilling is really easy to poke holes in.

  5. 1 month ago
    Anonymous

    https://lore.kernel.org/lkml/[email protected]/
    Lmao, the dude literally bumped an email trail from 2018 that failed to gain any traction and is begging for C++ to be allowed.

    • 1 month ago
      Anonymous

      >nvidia shows up

  6. 1 month ago
    Anonymous

    Rust is so comfy bros never going back

    • 1 month ago
      Anonymous

      Some things cannot be added back after being subtracted. Many such cases.

      • 1 month ago
        Anonymous

        I’ve still got my dick anon

  7. 1 month ago
    Anonymous

    ceniles are panicking by the fact that their dogshit language is fading.

  8. 1 month ago
    Anonymous

    >Note that no one in their sane mind would expect to use all the features of C++
    he really is trying to do that marketing speech...
    Linus can see through their bullshit but the trannies might have already tortured him too much.

    • 1 month ago
      Anonymous

      I also gulped on that sentence. It’s such an absolute statement that puts shame on hypothetical people that suddenly implies the opposite of what he is literally saying.

      What the whole thread just really means: the Rust kernel layer is a gigantic fail since C is not compatible with Rust

      My guess is that the guy will turn into a woman

  9. 1 month ago
    Anonymous

    If all of the advocacy from Rust trannies ultimately resorts in Linus favouring modern C++ in the kernel over their crude mockery of a language, ohhh I'll be laughing, might even buy a few more monitors and rest my feet on the desk.

    • 1 month ago
      Anonymous

      C++ is still an improvement over C.

      • 1 month ago
        Anonymous

        Yes, because C programs of large size end up trying to approximate something which is better thought out in C++ anyways. I've heard that the Linux kernel actually has a lot of object oriented design.

        • 1 month ago
          Anonymous

          >I've heard
          the absolute state of sepples zealots

        • 1 month ago
          Anonymous

          >I've heard that the Linux kernel actually has a lot of object oriented design.
          In the sense that early OOP is just normal procedural things like functions that operate on a struct, structs of function pointers used as interfaces, and struct pointer aliasing to implement something approximating inheritance, but with better syntax, yes, but so do all C programs. C railroads you into that.
          It doesn't have inheritance hierarchies or adhere to SOLID.

          • 1 month ago
            Anonymous

            so does GGML amusingly, they wrote it so much like a C++ library it makes me think the devs must have wanted to use C++ but a C lib was just easier to consume for 3rd party languages.

            But when C++ modules are implemented completely (there is a standard/spec document that details the type information contained in the .lib files, and possibly someday DLL files with ABI versioning restrictions)--it will be possible for other languages like python to import the trimmed-down exported C++ types (classes even, presumably, if the consuming language supports them) without any wrapper written manually at all.

            No need to attach C++ header files to your .lib because they contain the type information in a form that is simple and easy to use. Yes this does mean there will be restrictions on exported types in a C++ library project.

          • 1 month ago
            Anonymous

            >GGML
            At a glance it looks pretty norm
            Have a look at GLib and Gtk

          • 1 month ago
            Anonymous

            >SOLID

            even SOLID is old and busted these days, get with the CUPID/whatever else is amalgamating at the moment. At the end of the day, we just don't make everything a class on purpose in modern C++ and inheritance is bad. The base class of evil, even.

          • 1 month ago
            Anonymous

            I have a mess of bare metal C that runs on a riscv fpga soft core. I will port this code to c++ in a heartbeat if I can. To have templates in deep embedded code would be great.

            is bad

            wat. Its fine. just don't overuse OO.

  10. 1 month ago
    Anonymous

    > redhat
    > "oh boy I surely hope they are not ill-intentioned"
    > clueless.png

  11. 1 month ago
    Anonymous

    Amazing news! If C++ is good enough for the F-35, surely it's good enough for the kernel!

    • 1 month ago
      Anonymous

      Ada to C++ is worse. C to C++ is an improvement.

      • 1 month ago
        Anonymous

        C++ is far more popular than Ada, that has to mean something

        • 1 month ago
          Anonymous

          It means most software doesn't have to be reliable.

          • 1 month ago
            Anonymous

            C++ is reliable in the hands of anyone competent enough to use it correctly

          • 1 month ago
            Anonymous

            Most software doesn't need five nines reliability

          • 1 month ago
            Anonymous

            It's this kind of attitude that has turned everything into shit.

            They should covert it to Zig

            Zig, while not even at 1.0 yet, is already a better language than Rust.

          • 1 month ago
            Anonymous

            bitcoin core is in C++

        • 1 month ago
          Anonymous

          >argumentum ad populum anironically

  12. 1 month ago
    Anonymous

    RUST TROONS BTFO

  13. 1 month ago
    Anonymous

    Without exceptions it's not C++, with exceptions it's not usable.

    • 1 month ago
      Anonymous

      >Without exceptions it's not C++

      without exceptions you have error codes or calling std::terminate like R*st does.

      Some of the std lib is built with error codes in mind, not much, however. Whatever they do, it's probably going to be better than fricking GetLastError() or whatever other out-of-band error handling mechanisms they have in the C kernel.

      • 1 month ago
        Anonymous

        >without exceptions you have error codes or calling std::terminate like R*st does.
        You can do that, in a language which is not C++.

        It superficially looks like it, but it ain't. If you link it with proper C++ code it will break exceptions trying to pass through. It's not a subset, it's a dialect.

        • 1 month ago
          Anonymous

          And the existing code is written in a C dialect and yet
          This objection seems kind of uninteresting

    • 1 month ago
      Anonymous

      I don't think anyone except boost developers actually like exceptions.
      exceptions are pretty easy to avoid, and every major project disables them.

      • 1 month ago
        Anonymous

        most of the std lib has noexcept versions of almost everything, so it's not a major issue unless you absolutely must handle that error even though you raised an exception which just called std::terminate

      • 1 month ago
        Anonymous

        and I will say my day job is working on a GDK that does have exceptions enabled, they probably just get logged before the process dies and is restarted though.

  14. 1 month ago
    Anonymous

    >assembly
    >C
    Very nice. They should keep using assembly, keeps brainlets out.

    • 1 month ago
      Anonymous

      they probably mean inline assembly mostly, though idk.

  15. 1 month ago
    Anonymous

    Torvalds hated C++ last I heard so I doubt this will go anywhere, but I will watch the inevitable flamewar with glee

    • 1 month ago
      Anonymous

      He only hated it because at the time it was new and popular with subpar programmers
      These days C++ programmers are so competent with debugging, static analysis, and better error messages in the compiler that there's no reasoSegmentation fault (core dumped)

      • 1 month ago
        Anonymous

        kek

        OOP is the bad part of C++ in reality though

      • 1 month ago
        Anonymous

        The timing is perfect with systemd's bluescreen though. Just put the catchall block in the bluescreen handler.

      • 1 month ago
        Anonymous

        That was a pretty goodIn post included from
        from

        Torvalds hated C++ last I heard so I doubt this will go anywhere, but I will watch the inevitable flamewar with glee

        from

        https://i.imgur.com/X7ptQmz.png

        C++ is the future!
        https://www.phoronix.com/news/CPP-Linux-Kernel-2024-Discuss
        >"Now, "why not Rust"? First of all, Rust uses a different (often, in my opinion, gratuitously so) syntax, and not only would all the kernel developers need to become intimately familiar to the level of getting the same kind of "feel" as we have for C, but converting C code to Rust isn't something that can be done piecemeal, whereas with some cleanups the existing C code can be compiled as C++.
        >However, I find that I disagree with some of David's conclusions; in fact I believe David is unnecessarily *pessimistic* at least given modern C++.
        >Note that no one in their sane mind would expect to use all the features of C++. Just like we have "kernel C" (currently a subset of C11 with a relatively large set of allowed compiler-specific extensions) we would have "kernel C++", which I would suggest to be a strictly defined subset of C++20 combined with a similar set of compiler extensions.) I realize C++20 compiler support is still very new for obvious reasons, so at least some of this is forward looking."
        @
        rusttrons BTFO!

        :
        error: no match for 'template post &operator good(const joke &)'
        note: in instantiation of 'post' [aka 'struct comment_post']
        note: instantiated from here
        note: candidates are:
        note: 'template post &operator good(const point &)'
        note: 'template post &operator good(const post &)' [aka 'template post &operator good(const struct comment_post &)'

  16. 1 month ago
    Anonymous

    Linux should be rewritten in hare

    • 1 month ago
      Anonymous

      buy an ad Drew

    • 1 month ago
      Anonymous

      drew would disagree
      >In my opinion, Hare does not belong in the Linux kernel
      http://webcache.googleusercontent.com/search?q=cache%3Adrewdevault.com%2F2022%2F10%2F03%2FDoes-Rust-belong-in-Linux.html

  17. 1 month ago
    Anonymous

    Both c++ and rust are garbage, but tust is even worse than c++, be cuse compile times are even slower

  18. 1 month ago
    Anonymous

    Why don't they jusft fork it?
    >Red Hat engineer
    nvm

  19. 1 month ago
    Anonymous

    They should covert it to Zig

  20. 1 month ago
    Anonymous

    I think Rust has no business in kernel level programming and i say this as someone who thinks the language does have a usecase in some scenarios.

  21. 1 month ago
    Anonymous

    The key point here is to have a "kernel C++" subset. C++ is fine. I'd personally be more of an advocate for C++ and never consider C if they replaced name mangling, and stopped trying to support every paradigm.

  22. 1 month ago
    Anonymous

    #NotHappening
    Sepples are delusional

  23. 1 month ago
    Anonymous

    >Linux kernel about to get carpet bombed with C++ templates
    unless linus will gatekeep templates away, it's literally over for linux. frick debugging your "modern c++".

    • 1 month ago
      Anonymous

      debugging function pointers is worse than anything modern C++ has to offer except inheritance (runtime) polymorphism

  24. 1 month ago
    Anonymous

    Kek, Wait till Linus dies, you won't get luck adding C++ to the kernel before that.

  25. 1 month ago
    Anonymous

    Why is it always Red Hat? Can't they just leave shit alone? Why are there no linux forks? There's several BSDs, how long until someone fixes this shit.

  26. 1 month ago
    Anonymous

    I will back a "kernel c++" so long as it removes inheritance and exceptions

  27. 1 month ago
    Anonymous

    >Rust uses a different (often, in my opinion, gratuitously so) syntax
    this is actually my biggest problem with rust
    algol and ml syntax don't mix
    they're both good on their own
    but mix them and it's frick ugly

  28. 1 month ago
    Anonymous

    I wonder what aspects of C++14 he means when he says they make metaprogramming better. I don't remember any significant changes on that front until C++17, where they added if-constexpr. but yeah, C++20 with concepts would be ideal.
    anyway, Torvalds will torpedo this initiative to shit, but it's good to know there are some sane kernel devs out there.

  29. 1 month ago
    Anonymous

    Actually you can enforce a subset. It's not about the programming language, but how it's used. C++11

  30. 1 month ago
    Anonymous

    What? Months ago trannies were gloating Linux was trans and now this? Based, I hope thevmain guy in Linux Kernel development is a straight white man, lol.

    • 1 month ago
      Anonymous

      What?

Comments are closed.