💾 Archived View for gem.hexandcounter.org › gemlog › posts › hgamedev.gmi captured on 2024-12-17 at 10:10:31. Gemini links have been rewritten to link to archived content
-=-=-=-=-=-=-
Some comments on this:
gemini://gemini.ctrl-c.club/~lettuce/hobby-games.gmi
One idea I had is to implement solo boardgames on a computer. That solves all the AI problems! Guess that is basically the opposite of the approach in the linked post. In particular I was always a fan of solo (or coop) games that build the opposition into the rules of the game, in a way that means there really isn't any AI or bot players to worry about. I think more digital games could do that, in general. I never felt like a game was improved by trying to pretend there is a real opponent somehow outside of the game playing against me, rather than me playing against the game itself.
I know several 19th century or early 20th century games that could be fun to implement. Using such old games also means that the games could be distributed without worrying about copyrights etc. But those are not solo games and would require AI code to be playable (I am no fan of multiplayer games in general). Often quite non-trivial AI I suspect.
Once I made a digital version of a boardgame, and had to add some AI code. It was possible in that game to make a reasonable implementation by enumerating all possible moves, assigning a score to each, adding some randomness (to make the AI less predictable), and then pick the move with the highest score. Adding some long-term planning would probably be necessary for most games and that obviously makes things more complex. Games focusing more on the short-term tactics over longer term strategy can probably often be solved using some simple trick like this, but not for more strategic games.
One possible loophole I thought of when it comes to distributing implementations of copyrighted games (but I am not a lawyer) is to build something on top of freely available boardgame modules for virtual tabletops. If the publisher has given permission to distribute a game module for something like VASSAL or ZunTzu or Cyberboard, then it should be reasonably legal to distribute a game-specific player for that module? So players would still have to download the module file from somewhere else, but instead of playing in the application it was designed for they would load it into some game-specific player that provided a (subjectively) nicer GUI and maybe AI. I thought about making front-ends for a few of my favorite solo boardgames that way. It would make it difficult to use something like Pico-8 or TIC-80 or anything in-browser probably, but Löve 2D or Godot or Raylib or most other game frameworks should be able to load some graphics from a downloaded file at run-time.
There was a series of posts on BGG a while ago by someone showing how they liked to make, for their own personal hobby use, implementations of boardgames in Excel (or maybe it was Google Sheets?) and it was fun to follow. Much work could have been saved by just dragging some images into Godot and adding a few scripts there, or just make a VASSAL module, instead of all the tricks that went into specifying actions as buttons in a spreadsheet, and of course the result relied on all sorts of strange semi-automated interactions because of how a spreadsheet obviously is not meant for playing games. But as a hobby it did not seem like a boring activity.
tags: #games #programming