• rockerface 🇺🇦English
    191•1 year ago

    As a software developer, the less ambiguous your notation is, the better it is for everyone involved. Not only will I use brackets, I’ll split my expression into multiple rows and use tabs to make it as readable as humanly possible. And maybe throw a comment or 2 if there’s still some black magic involved

    • @dustyData@lemmy.worldEnglish
      57•1 year ago

      As a professor said, most programming languages don’t care about readability and whitespace. But we care because humans need it to parse meaning. Thus, write code for people, not for the machine. Always assume that someone with no knowledge of the context will have to debug it, and be kind to them. Because that someone might be you in six months when you have completely forgotten how the code works.

      • @zalgotext@sh.itjust.worksEnglish
        16•1 year ago

        Exactly. You read code way more times than you write it, so it makes all the sense in the world to prioritize readability.

      • @snowsuit2654@lemmy.blahaj.zoneEnglish
        2•1 year ago

        Yeah I totally agree. You can minimize and optimize as part of your build procedure/compilation but the source code should be as readable as possible for humans.

      • rockerface 🇺🇦English
        6•1 year ago

        Yep, if you’re writing code for a machine, just do it in binary to save compilation time (/s just in case). Also, you in six months will indeed be someone with no knowledge of the context. And every piece of code you think you write for one-time use is guaranteed to be reused every day for the next 5 years

    • @MotoAsh@lemmy.worldEnglish
      9•1 year ago

      I genuinely hate being human for this stuff. So many things have such crazy computational shortcuts, it’s sometimes difficult to remember which part represents reality. Outside of the realm of math, where “imaginary” numbers are still a touch of enigma to me, so many algorithms are based on general assumptions about reality or the specific task, that the programmatic approach NEVER encapsulates the full scope of the problem.

      As in, sometimes if you know EXACTLY how a tool works, you might still have no idea about the significance of that tool. Even in a universe where no one is lazy, and everyone wants to know “why?”, the answers are NOT forthcoming.

    • @Quetzalcutlass@lemmy.worldEnglish
      49•1 year ago

      I had someone submit a pull request recently that, in addition to their actual changes, also removed every single parenthesis that wasn’t strictly necessary in a file full of 3D math functions. I know it was probably the fault of an autoformatter they used, but I was still the most offended I’ve ever been at a pull request.

    • steboEnglish
      0•1 year ago

      Ok but that’s unrelated to putting some numbers and operations in a calculator. No one is going to proofread that. If anything, you simply calculate it again.

  • @MotoAsh@lemmy.worldEnglish
    9•1 year ago

    The underlying truth of this joke is: Programming syntax is less confusing than mathematical syntax. There are genuinely ambiguous layouts of syntax in math (to a human reader that hasn’t internalized PEMDAS, anyways) whereas you get a compilation error if ANYTHING is ambiguous in programming. (yes, I am WELL aware of the frustrations of runtime errors)

    • @dejected_warp_core@lemmy.worldEnglish
      1•1 year ago

      Also: sometimes, a mathematician just has to invent some concept or syntax to convey something unconventional. The specific use of subscript/superscript, whatever ‘phi’ is being used for, etc. on whatever paper you’re reading doesn’t have to correlate to how other work uses the same concepts. It’s bad form, but sometimes its needed, and if useful enough is added to the general canon of what we call “math”. Meanwhile, you can encapsulate and obfuscate things in software, sure, but you can always get down to the bedrock of what the language supports; there’s no inventing anything new.