AI-assisted development is bringing old games to platforms we'd never have thought possible. That's exciting, but the growing flood of unfinished ports is beginning to raise an uncomfortable question: how many of these games are actually worth playing?
  • Kiernian@lemmy.worldEnglish
    2·
    8 hours ago

    I read “port slop” and said “yeah, it’s called UDP, but that’s not new, why is there an article about it? click…oh.”

    /facepalm

    I need to stop browsing before I’ve had caffeine.

    • PerfectDark@lemmy.worldOPEnglish
      1·
      7 hours ago

      I hope you still enjoyed the read, even if it ended up being completely different to what your pre-caffeine brain saw!!!

  • ATPA9@feddit.orgEnglish
    1·
    8 hours ago

    I saw a video about a vibecoded ratchet and clank port. It BARELY managed to get over 100 fps and dropped to 60 when near water on a 3070.

    A game from 2001 should not run that badly on modern hardware, even with some visual enhancements.

  • PerfectDark@lemmy.worldOPEnglish
    61·
    1 day ago

    Some of the ports we’re seeing in the retro handheld community lately have been great, and I love seeing developers push these little devices beyond what they were ever intended to do. But I’ve definitely noticed what feels like a growing amount of ‘port-slop’, with new projects appearing seemingly every few hours, many of them feeling more like quick experiments than anything particularly worthwhile.

    I’m not against the experimentation at all, but I do wonder if we’re reaching a point where quantity is starting to overshadow quality. Anyway, I put together a short opinion piece about it, just my own thoughts on something I’ve been noticing lately, especially here with my favorite old system: the PSP.

  • JensSpahnpasta@feddit.orgEnglish
    47·
    21 hours ago

    People like to hate on everything AI, but where is the problem here? You are getting ports of games. Which is awesome. The project might be abandoned by the dev sometime in the future, but that is also happening with every other hobby project out there. And those ports are also not taking that many tokens to produce, so even the energy consumption argument is not really working here. Someone going on a ski trip will produce much, much more CO2 than someone creating a port from home.

    • MimicJar@lemmy.worldEnglish
      2·
      8 hours ago

      but where is the problem here? You are getting ports of games.

      One of the questions the author asks is exactly that, are you really getting ports of games?

      how many people are actually playing these AI-assisted ports? Not watching demonstrations, not congratulating their creators, but installing them, putting significant time into them and enjoying the games themselves.

      The author doesn’t have an answer, but it is a question that requires further investigation.

      If instead of AI we had a magical machine that gave you a bike. That’s great! You can do all sorts of things with your bike. So you hop on down to the movie theater, but the wheels fall off half way. You can’t fix the bike, so you walk home and miss the movie.

      Now you’re right that any hobby project might get abandoned. But there were also generally only a few hobby projects. They had different reputations and track records.

      Back to our bike example. Now we have hundreds of thousands of different bikes being printed from our magical machine. How do we pick the right one? What if all the bikes fail? What if there are a few that are good? How do we find the good ones?

      The answer here might just be waiting to see how this all shakes out. Everyone has an oven at home, and everyone can make cookies. It’s great that everyone can do that. But not everyone can be a baker. Or I’m wrong. Maybe with AI everyone can be a baker. That is one of the arguments.

      • JensSpahnpasta@feddit.orgEnglish
        11·
        9 hours ago

        Yeah and that is important: You might prefere that. But there are also people who do not speak or read English or other foreign languages and whose mother tongue is kind of obscure with not that many native speakers. For them this book they want to read is not available in a human-translated version. Machine translation gives them the ability to access the book. Nothing to complain about here.

        In this case: There seem to be not that many people around with the skills and the interest to do those ports manually. So the deal here is “AI version” vs. “Nothing”. You are free to chose “nothing”, but I really do not understand why you would complain here.

        • Lauchmelder@feddit.orgEnglish
          1·
          7 hours ago

          It’s not like there’s nobody around to decompile and subsequently port games. Up until a short while ago there was a community of decompilers that ultimately shut down because there were too many people forking ongoing projects just to shit out a slop port of the things they were working on. It’s not as simple as “AI porting games to obscure platforms”, it’s also a lot of “AI taking existing work, slopping it to the finish line and then flooding the market”. The market is flooded with machine translations of the same book.

          Check out https://decomp.dev/

          Machine translation gives them access to the book; does that mean we should create dozens of machine translations of every book out there, for every possible language?