Customization for creators using the MUD engine

[Jasrags]

While working on Use accessors for mob/player stats and reflecting on the related discussions, we should take a step back and consider how we want to support customization for creators building on top of the MUD engine.

How do you think we should support importing custom worlds and extensions without requiring direct modifications to the core codebase? Are there approaches you've seen work well (or poorly) in other engines for things like custom game data, rulesets, or modular logic?

One model that comes to mind is Ranvier's bundle system, where developers can create self-contained modules that define areas, commands, behaviors, and more without touching the core engine. This allows for a clean separation between engine and game content, making it easier to manage and share customizations.

[Volte6]

We have been developing out a module system. See the /modules/ folder

[Jasrags]

I was looking into that as well. I am thinking about a high level discussion around what systems should be considered part of the "base" modules, and how creators might integrate their own modules or areas. Should we expect creators to be able to introduce entirely new systems, like a custom spell framework, without needing to modify the core engine?

[Volte6]

Insofar as we can support it, yes. It needs to be somewhat intuitive and each feature we add module support for requires some customization, so it needs care/consideration... it's not really an all-at-once sort of thing.

Some examples of what we support right now: https://github.com/GoMudEngine/GoMud/tree/master/modules#things-modules-can-do

I think we'll keep building on it, and refactoring pieces as needed. We just need to be smart about it.

The more of the "base game" we move to modules, the more customizable it becomes. Ideally long term we'll move more and more of it to modules that can be removed or replaced, but certain cross-dependencies can make that tricky to do well.

[Jasrags]

Agree, the slow migration of core systems into modules makes sense to me and I also agree it will be complex :)

[Volte6]

NOTE: The migration of commands and features into modules enables us to discover new functionality/support to add to the modules/plugins system in a controlled and increment manner.

Ideally no changes would be implemented in modules/plugins without it being exercised in an actual in-game module, for dogfood-eating reasons as well as being a place in code we can point to for programmers looking to implement their own modules.