Would you be able to share some more details about the roadmap for 2024?
I know it was mentioned in a blog post that this will be the year of 3d, which is really exciting. I feel there is a spot for a leaner engine in the 3d space, not chasing render fidelity and realism, and Defold could feel that area impressively well.
Also, I’m not sure if I’m the only one, but I really enjoyed the get together we had a few years back via google meet. It’s a way to see people involved in the community and express our appreciation for the work that is being done.
Also very keen on 3D features. If you could painlessly mix Defold’s already really good 2D with the new 3D stuff that’s even better. Like displaying Spine or Rive content on a quad in 3D space for example.
“This year will be the year when 3D games in Defold will really take off. We’ve now reached a point where Defold has the required building blocks to build visually impressive 3D games. We still miss a few things (for instance compute shaders, instancing and GPU skinning) but these will be added in 2024 together with new 3D examples to help developers get started.”
Compute shaders, instancing and GPU skinning are obviously tasks that will fall upon us in the dev team, but then we hope you guys will turn all of the new 3D and rendering features into nice looking games! Another thing we’ve talked about is animation blending.
Another thing which will likely happen this year is that we start to move the 2d and 3d physics engines into their own extensions. Initially as internal extensions, but eventually they might become external extensions. Splitting up the physics into a 2d and 3d api will allow us to more rapidly add physics engine specific features and terminology into each api.
In the blog post we also mentioned HTML5 games in Defold. The platform is very important to us and we will continue to improve upon the stability and performance of web exports. Possibly also WebGPU.
Improved editor scripting will be one of those big impact things which we must start to deliver on this year. I’m not sure how far we’ll reach this year, but we need to see things happen this year. The idea is that you should be able to use Lua to script certain aspects of the editor and build your own tools.
We enjoyed it too. We need to find a format for these kinds of online events which is easy to maintain over time. What would you guys like to see during some kind of interactive online events?
The #defold-stage channel on the Discord server is built for live events like the one from 2021. It allows everyone to join and listen / watch, but nobody can speak unless the host allows them to. Perhaps look into utilizing this?
I think 3D even more than 2D requires a solid editor workflow, but since everyone is already invested in their own dcc (Blender, Maya, etc) I would love to have an easy way to import from those with or without a middle man format. FBX and Blend files come to mind for sure.
My wish is that beside of Defold engine development, the Defold editor is being evolved as well - especially allowing some kind of plug-in or modification in order to make Defold a plug-and-play solution, but suited for different needs.
I imagine teams being able to include needed tools inside Defold, like starting with more advanced tilemaps features or some sprite or sound editing inside or even lacking now physics joints editor to some more specific built-in “editors” like level editor that includes game specific toolset, specialized Lua tables editor with data that is presented in a way that developers want it to, some kind of narrative design tooling, etc.
Yes, that might be a good solution since the community is already there on Discord. The disadvantage is that it seems like there’s no built in functionality to record the stage channel? I’ve looked at Google Meet because it has a Stream (and record) to YouTube feature. The Zoom solution we had last time was too complicated.
i wish to see in defold a bit more productive editor like allowing drag and drop and also when choosing input instead of searching in that long list why not just detect input also to make prototyping easier make camera follow the player it’S assigned to without the need of code but also code for more custom feel and like @Pawel said why not add something like in godot to make full editor tools with lua like add new component wich will make the engine more modular and also you will have more feature from the community that can be added to the engine in future like godot did for the heightmap node plugin
As im interested in doing a 3D game but not wanting to switch to other game engines, im really happy to read that you guys want to make 3D this years topic.
I just want to know if ‘This year will be the year when 3D games in Defold will really take off’ is more geared towards mobile or will also apply to desktop? What’s your vision?
It’s understandable that it won’t have the graphic quality of Unity or Unreal (atleast not in 2024 ), but I would still like it to look good. Should I switch or grow with the engine and hope it turns out well? Difficult decision. My expectation for looking good would be something like Super Mario Bros. Wonder
Well, this year we’ll hopefully have time to add a few more capabilities:
Instancing
Animation retargeting
Animation blending
Improved glTF importer
Custom collision shapes
When we have the above in place I don’t see anything that really prevents a motivated Defold developer from creating basically any type of good looking 3D game. We should preferably also provide a few sample projects to help new developers get started. Also remember that anything we do in Defold will follow the principle of providing low-level building blocks rather than high level features, which means that you as a developer will have to do a bit more work yourself.
Do you plan to keep using Java ( jvm ) in Engine ?
Is JVM or Java that important in Defold ?
I have no technical knowledge but i have experienced Java is bit heavy and is memory hog and cpu drainer. While c# is memory Hog.
I don’t know about Mac or linux or very high end Windows pc. But on Mid end laptops, Heavy-ness of the software can be felt. It drains my cpu on 60-80% when 6 tabs opened and game launched in editor.
Is there a way to make engine completely with C++,c like efficient languages ?
I don’t know where to write what to add in the engine.
But I think, it’s good idea to add “Cinematic mode” like Godot or Unreal Engine.
Frame per frame screenshort saving (png, jpg, gif, avc, mkv etc).
It is good for procedural animations.
Also, I think it’s needed “Drug and Drop” option for file manager: “Directly from floder to an editor”.
Some kind of integration I guess, but what precisely?
), but I would still like it to look good. Should I switch or grow with the engine and hope it turns out well? Difficult decision. My expectation for looking good would be something like