• shape_warrior_t@programming.dev
      link
      fedilink
      English
      arrow-up
      32
      ·
      3 months ago

      I would certainly rather see this than {isAdmin: bool; isLoggedIn: bool}. With boolean | null, at least illegal states are unrepresentable… even if the legal states are represented in an… interesting way.

        • shape_warrior_t@programming.dev
          link
          fedilink
          English
          arrow-up
          9
          ·
          3 months ago

          I was thinking of the three legal states as:

          • not logged in (null or {isAdmin: false, isLoggedIn: false})
          • logged in as non-admin (false or {isAdmin: false, isLoggedIn: true})
          • logged in as admin (true or {isAdmin: true, isLoggedIn: true})

          which leaves {isAdmin: true, isLoggedIn: false} as an invalid, nonsensical state. (How would you know the user’s an admin if they’re not logged in?) Of course, in a different context, all four states could potentially be distinctly meaningful.

    • grrgyle@slrpnk.net
      link
      fedilink
      arrow-up
      8
      ·
      3 months ago

      The variable name is 90% why this is so unreasonable. Code is for humans to read, so names matter.

    • Drewmeister@lemmy.world
      link
      fedilink
      arrow-up
      8
      ·
      3 months ago

      E: omg forget my whole comment. I agree with you that the name sucks.


      I mostly don’t like that role is typically an intuitive name, and now suddenly it means something I wouldn’t expect. Why add confusion to your code? I don’t always remember what I meant week to week, much less if someone else wrote it.

      • dohpaz42@lemmy.world
        link
        fedilink
        English
        arrow-up
        7
        ·
        3 months ago

        If I had a nickel for every time that happened to me, I’d still be poor, but at least I’d have several nickels. 😁

    • normalexit@lemmy.world
      link
      fedilink
      arrow-up
      3
      ·
      3 months ago

      Product manager: “I want a new role for users that can only do x,y,z”

      Developer: “uh… yeah. About that… Give me a few days.”

      • grrgyle@slrpnk.net
        link
        fedilink
        arrow-up
        5
        ·
        3 months ago

        Hmmm I need a datatype with three states… Should I use a named enum? No, no that’s too obvious…

    • orgrinrt@lemmy.world
      link
      fedilink
      arrow-up
      1
      arrow-down
      2
      ·
      3 months ago

      Yeah let’s use a union of a boolean and null to represent role, something that inherently represents more than two (…or three, I guess) different values, as opposed to something like an integer.

      Even if the name is clearly misleading in this specific case, the entire choice of using a bool here is just bad because it’s almost guaranteed you’re going to expand on that in future and then you’ll just have to entirely rewrite the logic because it simply can’t accommodate more than two values (or three with the null union… 🙈), while it gives absolute zero benefits over using something more reasonable like an integer to represent the roles, or in this case, admin, not-admin and guest. Even if you’ll end up with just admin, non-admin and guest, the integer would still work great with no disadvantages in terms of amount of code or whatever. Just increased legibility and semantical accuracy.

      Not to mention that there’s zero reason to combine the state of being logged in and the role in which you’re logged in in one variable… those are two different things. They will remain two different things in future too…

      I mean they’re already chaining elseifs (basically matching/switching, while doing it in an inefficient way to boot 🥴) as though there were an n amount of possible states. Why not just make it make sense from the start instead of whatever the hell this is?