Godot Engine at gamescom 2023
Table of Contents
- Attending gamescom
- Meeting the Godot team and W4 Games
- Talking points with the Godot team
- Conclusion
Attending gamescom
Third time is a charm. This year was my third attempt to attend the gamescom fair in Cologne, Germany. The previous two attempts – around 2018 and 2019 – did not work out, since I did not want to wade through the massive crowds on the weekends.
This time I managed – after some hassle – to get myself trade-visitor tickets and attend gamescom during the business days, which have a much reduced people-density.
Meeting the Godot team and W4 Games
On two afternoons I had the chance to talk extensively with Emilio Coppola, Clay John and very briefly with Rémi Verschelde. I also caught a glimpse of Nathan Lovato of GDQuest fame, but did not get a change to chat with him.
Talking points with the Godot team
How to get OIT into Godot?
From the header video you can see that I am eager to have order-independent transparency (aka OIT) in Godot Engine. Since this feature is not on Godot’s development roadmap, their development resources are limited (as enthusiastic as they are, they are realistic… they are not Epic or Unity folks) and OIT is a niche high-end rendering feature, implementing this in Godot myself is the only way to get this in a reasonable time frame.
With Clay John I talked about the best approach to hook it into Godot. Since GDExtension does not expose the inner workings of the engines rendering infrastructure, the best approach is to patch it into the renderer directly. I will focus on Vulkan first… just to get a working proof-of-concept in place from which I can improve and polish and perhaps later derive an OpenGL variation for completeness.
Another question was what requirements will need to be fulfilled by my patches to increase the likelihood of them being accepted upstream in a PR. The usual, expected and totally reasonable things like clean code, good rendering performance and opt-in default (read: don’t be enabled by default). But just a Vulkan-implementation first and an OpenGL one in a later pull-request is acceptable for the Godot team.
Will exporting become more modular?
The Godot Engine is pretty lean already… especially compared to Unreal, Unity, Unigine etc. But the exported binaries e.g. on x86-64 Linux or Windows are still pretty large. My small Rubik’s cube demo (Linux binary, Windows binary, screencast) is around 85 MBytes.
Upon me asking if there are any plans to allow the exporting of projects in a more modular fashion, I learned that such a feature is actually already planned and being worked on. So e.g. a pure 2D-platformer game would be able to leave out any 3D-renderer code, the 2D-UI toolkit could be skipped from exporting if not used at all, a 3D-fps shooter could skip all the 2D parts and so on.
How granular this will become I do not know. But this is terrific news… my demoscene-affiliated heart was full of joy learning about this feature! 🙂
Automatic testing and performance-regression safe-guards in Godot’s CI?
Developers will be developers… and care about automated build infrastructure, which helps maintain and improve quality and performance of a code-base.
While not in place on a broad scale yet, this is something that is greatly welcomed and appreciated by the Godot team. They have some things in the pipeline, but have to pick their battles carefully due to still limited resources.
If you live and breath QA and want to help out with the Godot engine project do not hesitate to lend a hand or two!
Are high-quality tech-demos and benchmarks planned?
The one thing that ‘sells’ a game-engine in the best possible way are jaw-dropping gorgeous tech-demos and benchmarks, which show off the features in breathtaking ways. I added my own 2-cents to that discussion with a small screencast a few months ago.
At the booth I learned, that W4 Games has such tech-demos in the works. But what exactly, how many and further details I did not figure out… guess it would ruin the surprise 🙂
The people around the Godot engine and W4 Games of course know, that such tech-demos and benchmarks are needed as a ‘promotional’ vehicle to further spread the word and awareness of the project.
Finally I asked them about where and how to hook up some very talented game-graphics designers I know with them and W4 Games. But details and plans are to early to share at the moment. This will take a few weeks I assume.
Conclusion
All in all a very informative, enlightening and pleasant occasion meeting some of the folks behind the Godot Engine project. This project has the potential to become the “blender of game-engines” in terms of scope, organization and feature-breadth in my opinion.
Hopefully I can support them and the Godot Engine by next year’s gamescom 2024 with my working OIT-implementation in the Vulkan-renderer (at least) and some booth-staffing-power, since I live only 70 km away from Cologne.