• mindbleach@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    59
    ·
    9 months ago

    I’ve done the math for how long it’d take to randomly guess the last several kilobytes until something checksummed correctly.

    I was not pleased with the answer.

    • holycrap@lemm.ee
      link
      fedilink
      English
      arrow-up
      27
      arrow-down
      1
      ·
      9 months ago

      That would put those crypto miners to better use at least

    • Catsrules@lemmy.ml
      link
      fedilink
      English
      arrow-up
      14
      ·
      9 months ago

      You know I never thought of that… but yeah that would be a good very very very very large number.

      Like throwing puzzle pieces in the air and getting it to land completed.

        • Trailblazing Braille Taser@lemmy.dbzer0.com
          link
          fedilink
          English
          arrow-up
          4
          ·
          9 months ago

          Haha agreed, if we’re talking about kilobytes of missing data brute forcing is intractable.

          There may be structure to exploit in the data format. E.g. if you’re recovering missing content from a book written in English, you can probably get away with enumerating only printable ASCII and 90% of the letters will be lowercase.

          But practically, I am unconvinced because the information density is pretty high on the kinds of things people like to torrent.

    • Laura@lemmy.ml
      link
      fedilink
      English
      arrow-up
      4
      ·
      9 months ago

      wait until you hear about collisions (missing more bits than your hash output length guarantees a collision on average)