5 ms·

Au contraire, mon ami: this feature is supported on all major compilers. ;)

dwulfen 11y ago
this is effectively bitwise math, so it's actually supported in any language offering the minus and binary complement prefix operators. Drop this into your address bar: javascript:var%20x=100;alert("x="+x+"\nx-1="+(~-x)+"\nx+1="+(-~x));
CamperBob2 11y ago
I'm pretty sure that C doesn't guarantee a two's-complement representation. Not so sure about C++. Either way, it makes for a cute novelty article, and a really terrible idea. Raymond should've saved it for the next April Fools' Day.
- dllthomas 11y ago
  It's supported by all major compilers, but some may have buggy implementations if you try to target a non-two's-complement architecture. Clearly they need to fix those bugs!
  - CamperBob2 11y ago
    Right, the tilde operator is supported, but the math trick in question only works on a twos-complement machine. So it's non-portable in the sense that the language spec doesn't guarantee a particular behavior.
    - dllthomas 11y ago
      Yup. If it wasn't clear, my comment above was a joke :-P
- kps 11y ago
  You got it. Both C (C11 §6.2.6.2) and C++ (C11 §3.9.1) permit two's complement, one's complement, and signed magnitude representation of integers. With one's complement, “−” ≡ “~” so “−~” and “~−” both do nothing. With signed magnitude, “−~” and “~−” both give the same result, which is the difference between the value and the maximum integer of the same sign.
  - tedajax 11y ago
    Is it unreasonable for me to want a C/++ standard that's just "What's mostly standard by all major compilers" so we can stop nitpicking about implementation details most people never encounter?
    - kps 11y ago
      At the time the first version of the C standard was being prepared, the practice of writing C that would only work with the ‘major compiler[s]’ had the nickname “All the world's a VAX”. Now consider the last time you used a VAX.
      - userbinator 11y ago
        The VAX is a little-endian, 8-bit bytes, 32-bit datapath, 2's complement machine. Consider the last time you used a machine with big endian, non-8-bit-bytes, and something other than 2's complement integer representation... At the time of the C standard, architectures were far more diverse; for better or worse, most of the more esoteric ones seem to have died off, leaving us with the ones that do happen to share many characteristics with a VAX.
        douche 11y ago
        Out of curiosity, are there any common architectures that are wildly different on those basises from x86/x64/ARM?
        userbinator 11y ago
        I can't think of any general-purpose architectures in common use which are really odd, but some DSPs have a 24-bit "byte". The microcontroller world is a bit more diverse too, especially with many of them being Harvard architectures (e.g. C compiles to 8051 with 24-bit big-endian pointers, and the various PICs have odd instruction-word sizes like 14 bits). However, all of those still use 2's-complement integers. According to Wikipedia a few existing mainframes use 1's complement (CDC 3000/6000, UNIVAC 1100/2200). I know the standard allows different bit representations, but practically speaking, today you're almost never going to find anything other than 2's complement integers and IEEE floating-point; and if you do happen to be working on one of those rare and unusual machines that don't, I think it'd be the least of your worries... even Linus says he doesn't care about Linux running on non-two's-complement or odd byte-size machines.

this is effectively bitwise math, so it's actually supported in any language offering the minus and binary complement prefix operators. Drop this into your address bar: javascript:var%20x=100;alert("x="+x+"\nx-1="+(~-x)+"\nx+1="+(-~x));

CamperBob2 11y ago

I'm pretty sure that C doesn't guarantee a two's-complement representation. Not so sure about C++. Either way, it makes for a cute novelty article, and a really terrible idea. Raymond should've saved it for the next April Fools' Day.

dllthomas 11y ago
It's supported by all major compilers, but some may have buggy implementations if you try to target a non-two's-complement architecture. Clearly they need to fix those bugs!
- CamperBob2 11y ago
  Right, the tilde operator is supported, but the math trick in question only works on a twos-complement machine. So it's non-portable in the sense that the language spec doesn't guarantee a particular behavior.
  - dllthomas 11y ago
    Yup. If it wasn't clear, my comment above was a joke :-P
kps 11y ago
You got it. Both C (C11 §6.2.6.2) and C++ (C11 §3.9.1) permit two's complement, one's complement, and signed magnitude representation of integers. With one's complement, “−” ≡ “~” so “−~” and “~−” both do nothing. With signed magnitude, “−~” and “~−” both give the same result, which is the difference between the value and the maximum integer of the same sign.
- tedajax 11y ago
  Is it unreasonable for me to want a C/++ standard that's just "What's mostly standard by all major compilers" so we can stop nitpicking about implementation details most people never encounter?
  - kps 11y ago
    At the time the first version of the C standard was being prepared, the practice of writing C that would only work with the ‘major compiler[s]’ had the nickname “All the world's a VAX”. Now consider the last time you used a VAX.
    - userbinator 11y ago
      The VAX is a little-endian, 8-bit bytes, 32-bit datapath, 2's complement machine. Consider the last time you used a machine with big endian, non-8-bit-bytes, and something other than 2's complement integer representation... At the time of the C standard, architectures were far more diverse; for better or worse, most of the more esoteric ones seem to have died off, leaving us with the ones that do happen to share many characteristics with a VAX.
      - douche 11y ago
        Out of curiosity, are there any common architectures that are wildly different on those basises from x86/x64/ARM?
        userbinator 11y ago
        I can't think of any general-purpose architectures in common use which are really odd, but some DSPs have a 24-bit "byte". The microcontroller world is a bit more diverse too, especially with many of them being Harvard architectures (e.g. C compiles to 8051 with 24-bit big-endian pointers, and the various PICs have odd instruction-word sizes like 14 bits). However, all of those still use 2's-complement integers. According to Wikipedia a few existing mainframes use 1's complement (CDC 3000/6000, UNIVAC 1100/2200). I know the standard allows different bit representations, but practically speaking, today you're almost never going to find anything other than 2's complement integers and IEEE floating-point; and if you do happen to be working on one of those rare and unusual machines that don't, I think it'd be the least of your worries... even Linus says he doesn't care about Linux running on non-two's-complement or odd byte-size machines.