• fullsquare@awful.systems
    link
    fedilink
    English
    arrow-up
    8
    ·
    1 day ago

    some people would tell you that we can simulate small bits of chemistry but it’s flat out wrong (i might be biased as i’ve wrangled for a year with computational chemists about results that don’t conform to reality) and even then errors are so large that’s it’s useless

    • mineralfellow@lemmy.world
      link
      fedilink
      English
      arrow-up
      9
      ·
      1 day ago

      I was involved with a project trying to simulate growth of a crystal cluster a couple of years ago. The guy doing the coding said it would be easy. It never worked and never came remotely close.

      • fullsquare@awful.systems
        link
        fedilink
        English
        arrow-up
        4
        ·
        23 hours ago

        in my case the size of the system was so small they didn’t have that excuse, yet they were only ever able to get correct results after experimental data was handed over to them, zero predictive power, useless

      • Zink@programming.dev
        link
        fedilink
        English
        arrow-up
        1
        ·
        1 day ago

        said it would be easy

        Ah, the innocence of inexperience. (Giving them the benefit of the doubt of course. )

        I work on old undocumented c/c++ spaghetti code for embedded systems. In multiple planning meetings I have gotten to use lines like “this looks like a single character change but testing it makes me really nervous” and have gotten zero pushback or raised eyebrows.

        It’s usually a few laughs and often another engineer or our manager will chime in to agree with me and describe some more of the context or whatever, lol.