• PixxlMan@lemmy.world
    link
    fedilink
    arrow-up
    4
    arrow-down
    1
    ·
    1 year ago

    To everyone commenting that you have to convert to binary to represent numbers because computers can’t deal with decimal number representations, this isn’t true! Floating point arithmetic could totally have been implemented with decimal numbers instead of binary. Computers have no problem with decimal numbers - integers exist. Binary based floating point numbers are perhaps a bit simpler, but they’re not a necessity. It just happens to be that floating point standards use binary.

      • PixxlMan@lemmy.world
        link
        fedilink
        arrow-up
        1
        ·
        1 year ago

        Wrong. Sounds like you think only fixed point/precision could be implemented in decimal. There’s nothing about floating point that would make it impossible to implement in decimal. In fact, it’s a common form of floating point. See C# “decimal” type docs.

        The beginning of the Wikipedia article on floating point also says this: “In practice, most floating-point systems use base two, though base ten (decimal floating point) is also common.” (https://en.m.wikipedia.org/wiki/Floating-point_arithmetic) Also check this out: https://en.m.wikipedia.org/wiki/Decimal_floating_point

        Everything in my comment applies to floating point. Not fixed point.

        • JackbyDev@programming.dev
          link
          fedilink
          English
          arrow-up
          1
          ·
          1 year ago

          I generally interpret “decimal” to mean “real numbers” in the context of computer science rather than “base 10 numbers”. But yes, of course you can implement floating point in base 10, that’s what scientific notation is!