Just take the string as bytes and hash it ffs

  • Björn Tantau@swg-empire.de
    link
    fedilink
    English
    arrow-up
    3
    ·
    23 days ago

    Take a string as bytes is bad with weird non-ASCII characters. Been there, been bitten in the ass by it.

    At least with e-mail clients different clients on different operating systems use different encoding by default for their passwords.

    With a router I could imagine different client apps following different standards.

    • expr@programming.dev
      link
      fedilink
      English
      arrow-up
      4
      ·
      23 days ago

      You don’t have to take arbitrary bytes. UTF-8 encoded strings are just fine and easily handled by libraries.

      • Björn Tantau@swg-empire.de
        link
        fedilink
        English
        arrow-up
        1
        ·
        23 days ago

        At least with e-mail clients different clients on different operating systems use different encoding by default for their passwords.

          • Björn Tantau@swg-empire.de
            link
            fedilink
            English
            arrow-up
            1
            ·
            23 days ago

            With a router I could imagine different client apps following different standards.

            Many routers can also be controlled via Telnet, which will also use different encodings depending on your OS.

      • tiredofsametab@fedia.io
        link
        fedilink
        arrow-up
        1
        ·
        23 days ago

        Y’all use UTF8? laughs in Japanese websites

        / can we please stop EUC-JP and SJIS and MS932 and all just switch to UTF8, please, Japan?!