Start moving stuff out of core and into offical GDExtension plugins.

[Shadowblitz16]

Describe the project you are working on

A mario game but this is not really relevant.

Describe the problem or limitation you are having in your project

It's not my problem but a problem with godot.

Basically issues are closed alot due people wanting to keep godot slim.
But this comes at a cost, features.

The point of this is to fix that.

Describe the feature / enhancement and how it helps to overcome the problem or limitation

The idea is to start moving stuff out of core and into official gdextension plugins.
This would not only allow dev's to add as many features as they want but it allows the user's to pick what they want to to include in their game.

This not only allows godot to include tons of features but it also significantly reduces the size of core.

Describe how your proposal will work, with code, pseudo-code, mock-ups, and/or diagrams

It would probably be split up into the follow...

• A plugin per node type
• A plugin per resource type
• A plugin per language

There would probably be the need to allow users to create their own plugin packs.

Editor features would probably be tied to the node type, so if you download a Tileset2D plugin you also get the editor window for editing tilesets.

If this enhancement will not be used often, can it be worked around with a few lines of script?

No.

Is there a reason why this should be core and not an add-on in the asset library?

It does exactly that, moves things into a asset library, and makes the godot engine more modular.

[AThousandShips]

Any specific features you have in mind?

Note that the commonly suggested example of GridMap was rejected previously so likely not going to be approved any time soon:

• Remove GridMap from core #5561

And any such move would be a major compatibility breakage so might not be realistic in 4.x

[Shadowblitz16]

No feature specifically, The idea was to slowly move core components into gdextension addons, and having core mainly be servers and gdextension and have everything else be modular.

[Shadowblitz16]

@Castro1709 what exactly don't you like?
This isn't about removing features it's about making Godot more modular and allowing users to choose what they want to include in the engine.

The idea is to still keep the features but move them to the asset store. they would still be officially supported

[Castro1709]

Hi, I usually don't elaborate on why I like or dislike a proposal, but now that you asked here's why.

First and my main reason is that GDExtension plugins are limited, meaning that there's stuff that you just can't do in GDextension. The Godot team will have to put a lot of work into that.

Second, when you install a GDExtension you have to put all the binary files into the project. And if you suggest that every node type and resource must be one extension, the though of all the stuff that is going to be in the project just sounds horrible imo. Not to mention the hassle of starting a new project and just start downloading all the stuff.

Third, if you dislike having features that you won't use and you want your editor to be more slim, you can compile the editor only with the stuff you want.

Lastly, I like how beginner friendly Godot is, and I just don't think this is a good approach for beginners.
And even if you invest the money and the time into fixing all that and making it as friendly as possible, I just think that we could have used all those resources into new features, fixing bugs or improving what we already have.

[dsnopek]

I'm generally in support of having less stuff in Godot itself, and moving more into GDExtension.

For example, it would theoretically be cool to not have any physics engine in Godot, and instead have a set of GDExtensions to choose from.

However, I think we'd need to first make some key improvements to GDExtension in order to provide a nice experience and workflow for users, for example:

• Global extensions so some features only need to be installed once, and not per-project
• Some mechanism to suggest extensions depending on what the user wants to do (ex if they try to use a physics node, prompt them to install a physics engine)
• Easier management of dependencies
• Easier management of updates

All of these things (except maybe point nr 2) are all things we wanted to improve anyway. Once those features exist, perhaps we can take a look at the Godot eco-system, and see if it makes sense to have a lighter core in a hypothetical Godot 5.