I’m curious how software can be created and evolve over time. I’m afraid that at some point, we’ll realize there are issues with the software we’re using that can only be remedied by massive changes or a complete rewrite.

Are there any instances of this happening? Where something is designed with a flaw that doesn’t get realized until much later, necessitating scrapping the whole thing and starting from scratch?

  • 🌘 Umbra Temporis 🌒@lemmy.dbzer0.com
    link
    fedilink
    English
    arrow-up
    3
    arrow-down
    5
    ·
    7 months ago

    What was stopping X just undergoing some gutting? I get it’s old and covered in dust and cobwebs but look, those can be cleaned off.

    “Scoop out the tumors, and put some science stuff in ya”, the company that produced that quote went on to develop the most advanced AGI in the world and macro-scale portable on-demand indestructible teleportation.

    • Possibly linux@lemmy.zip
      link
      fedilink
      English
      arrow-up
      3
      arrow-down
      2
      ·
      7 months ago

      Because we no longer have mainframes in computer labs. Each person now has there own machine.

      • atzanteol@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        3
        ·
        7 months ago

        And yet I play modern games on modern hardware with X just fine. It’s been extended a little bit since the 80s.

            • PrimalHero@kbin.social
              link
              fedilink
              arrow-up
              1
              ·
              7 months ago

              What part of 40 year old code that is so messed up that it’s not cleanable any more do you not understand.

              • atzanteol@sh.itjust.works
                link
                fedilink
                English
                arrow-up
                1
                ·
                7 months ago

                Of course it is. That’s propaganda. It’s hard, but possible. Probably not as hard as fighting Nvidia for 15 years either.

                • nyan@sh.itjust.works
                  link
                  fedilink
                  arrow-up
                  2
                  ·
                  7 months ago

                  Simply put, no one with the necessary skills has come forward and demonstrated the willingness to do the work. No programmer I’ve ever met enjoys wrestling with other people’s crufty old code. It isn’t fun, it isn’t creative, and it’s often an exercise in, “What the unholy fsck was whoever wrote this thinking, and where did I put the ‘Bang head here’ mousepad?” So getting volunteers to mop out the bilges only happens when someone really wants to keep a particular piece of software working. It’s actually more difficult than getting people to contribute to a new project.

                  So getting rid of X’s accumulated legacy cruft isn’t impossible, but I suspect someone would need to set up the “Clean up X” foundation and offer money for it to actually happen. (I’m no happier about that than you, by the way.)

                  • atzanteol@sh.itjust.works
                    link
                    fedilink
                    English
                    arrow-up
                    1
                    ·
                    7 months ago

                    Aye - there was definitely a lack of motivation there. It seems the X teams (XF86 and later Xorg) sorta ran out of juice at some point. Maybe Wayland has reinvigorated them since it’s much more exciting to write new code than fix old cruft.

                  • atzanteol@sh.itjust.works
                    link
                    fedilink
                    English
                    arrow-up
                    1
                    ·
                    edit-2
                    7 months ago

                    clean it up yourself if you love it so much

                    Are we 12?

                    You’re telling me a project that has taken 15 years and is just now getting decent nvidia support and which may someday allow applications to position their own windows is rousing success? Compared to a rework of an existing codebase? That has all the signs of a “we bit off more than we could chew”.

                    It’ll work, in the end. But 15 years of work on a migration from X11->“X12” or something would have likely been easier. Especially if they didn’t ignore nvidia along the way.