This commit is contained in:
Miloslav Ciz 2023-05-21 21:58:37 +02:00
parent 94c2c620d2
commit d8ac59a610

View file

@ -34,7 +34,7 @@ Compared to mainstream games, a LRS game shouldn't be a consumerist product, it
The paradigm of LRS gamedev differs from the mainstream gamedev just as the [Unix philosophy](unix_philosophy.md) differs from the [Window philosophy](windows_philosophy.md). While a mainstream game is a monolithic piece of software, designed to allow at best some simple, controlled and limited user modifications, a LRS game is designed with [forking](fork.md), wild [hacking](hacking.md), unpredictable abuse and code reuse in mind.
Let's take an example. A LRS game of a real-time 3D [RPG](rpg.md) genre may for example consist of several independent modules: the RPG library, the game code, the content and the [frontend](frontend.md). Yes, a mainstream game will consist of similar modules, however those modules will probably only exist for the internal organization of work and better testing, they won't be intended for real reuse or wild hacking. With the LRS RPG game it is implicitly assumed that someone else may take the 3D game and make it into a purely non-real-time [command line](cli.md) game just by replacing the frontend, in which case the rest of the code shouldn't be burdened by anything 3D-rendering related. The paradigm here should be similar to that existing in the world of computer [chess](chess.md) where there exist separate engines, graphical frontends, communication protocols, formats, software for running engine tournaments, analyzing games etc. [Roguelikes](roguelike.md) and the world of [quake](quake.md) engines show some of this modularity, though not in such a degree we would like to see -- LRS game modules may be completely separate projects and different processes communicating via text interfaces through [pipes](pipe.md), just as basic Unix tools do. We have to think about someone possibly taking our singleplayer RPG and make it into an MMORPG. Someone may even take the game and use it as a research tool for [machine learning](machine_learning.md) or as a VFX tool for making movies, and the game should be designed so as to make this as easy as possible -- the user interface should be very simple to be replaced by an [API](api.md) for computers. The game should allow easy creation of [tool assisted speedruns](tas.md), to record demos, to allow [scripting](script.md), modifying ingame variables, even creating [cheats](cheat.md) etc. And, importantly, **the game content is a module as well**, i.e. the whole RPG world, its lore and storyline is something that can be modified, forked, remixed, and the game creator should bear this in mind (see also [free universe](free_universe.md)).
Let's take an example. A LRS game of a real-time 3D [RPG](rpg.md) genre may for example consist of several independent modules: the RPG library, the game code, the content and the [frontend](frontend.md). Yes, a mainstream game will consist of similar modules, however those modules will probably only exist for the internal organization of work and better testing, they won't be intended for real reuse or wild hacking. With the LRS RPG game it is implicitly assumed that someone else may take the 3D game and make it into a purely non-real-time [command line](cli.md) game just by replacing the frontend, in which case the rest of the code shouldn't be burdened by anything 3D-rendering related. The paradigm here should be similar to that existing in the world of computer [chess](chess.md) where there exist separate engines, graphical frontends, communication protocols, formats, software for running engine tournaments, analyzing games etc. [Roguelikes](roguelike.md) and the world of [quake](quake.md) engines show some of this modularity, though not in such a degree we would like to see -- LRS game modules may be completely separate projects and different processes communicating via text interfaces through [pipes](pipe.md), just as basic Unix tools do. We have to think about someone possibly taking our singleplayer RPG and make it into an MMORPG. Someone may even take the game and use it as a research tool for [machine learning](machine_learning.md) or as a VFX tool for making movies, and the game should be designed so as to make this as easy as possible -- the user interface should be very simple to be replaced by an [API](api.md) for computers. The game should allow easy creation of [tool assisted speedruns](tas.md), to record demos, to allow [scripting](script.md) (i.e. manipulation by external programs, traditional in-game interpreted scripting may be absent, as mentioned previously), modifying ingame variables, even creating [cheats](cheat.md) etc. And, importantly, **the game content is a module as well**, i.e. the whole RPG world, its lore and storyline is something that can be modified, forked, remixed, and the game creator should bear this in mind (see also [free universe](free_universe.md)).
Of course, LRS games must NOT contain such shit as "[anti-cheating](anti_cheat.md) technology", [DRM](drm.md) etc. For our stance on cheating, see the article [about it](cheat.md).