Shaders can have visual editors and be easy to use…
It’s all about tracking shader handles, uniform handling, and things of this nature.
Heck; could have built in ai-shader-tracker system to make it even more easier in the long-term…
Parameters would be equations, variables and built in (editable) drop-down options; etc.
It’s really more to do with if an developer; who enjoys that mostly more and or less non-mind orientated way of coding and can handle that level of complexity at the same time.
Blender would be a great example of a shader editor; that works… we need that in all engines with a way more easier interface…
People don’t use shaders; why? :
because they don’t understand them (because the original developers of glsl, hlsl, SPIR-V compiler; don’t know what mind orientated coding is; more and or less) and or the engine is taking so much distraction time that they hardly ever see; shader usage… but for those who get past the wall of delayed time and see the beauty of shaders; there is a heck of a lot of potential.
I mean look at blender shaders they get used a lot.
You provide a supposed example of a shader editor which makes it easy to make shaders, then immediately disqualify it for having a way too difficult interface?
From experience: It’s often those who have less experience with shaders who ask for shader editors. Their expectations can be inflated and there is rarely a concrete idea for what they wish to create with the tool.
Photoshop doesn’t teach you about color theory or composition, so no matter how well you know the software, it is not going to aid you in any way whatsoever in creating good art if you yourself have no artistic talent. Similarly, a shader editor doesn’t teach you the underlying skills and principles necessary for authoring shaders, which means that by the time you are able to make a good shader you no longer need or want to use a simplified editor. Most useful shaders have been written already. It would be a very niche tool.
If I had to write a shader editor I would only add the most fundamental kinds of nodes but make sure they combine into a large matrix of possibilities. It wouldn’t be virtually limitless the way a proper shader editor is, but could have a good balance between ease of use and expressive possibility.
Think you missed the gdevelop amazing programmers step; who is capable of making those better user-interfaces; sorry for not clarifying that aspect as well.
Id make a shader category for each area of universal rendering and then give them adaptable/dynamic shaders.
Color Shader (with gradients and so on)
Math Shader (with equations and so on which can do matrix calculations)
Flow Shader (so can attach, detach at runtime and before runtime on different shaders ;etc )
There’s plenty more of course; but am not in the mood to write them out; sorry.
That’s not my concern. It’s not about creating a simplified shader editor. It’s, firstly, to have a space where you can code shaders, even in the rough with the code (which would give us the possibility of copying elsewhere and pasting into the shader code editor something like extensions), and give us the possibility of attaching them to objects. That’s all there is to it. Now, if you want to add the simplified shader editor, that’s a plus.
I suspect it’ll be asking for a truck load of issues with beginners.
GDevelop is a game engine that non-coders can use. Shader programming is not for beginners. Even Unity, with its visual shader, is not a walk in the park to get the right shader results.
If a lot of the issues on this forum are around basic logic flow and object usage, can you imagine the issues most noobs would have with a shader that wasn’t doing what they want or creating serious lag?
The only way I can see it remotely working is if the GDevelop devs make a simple interface for it, effectively limiting what the shaders can do.
That’s not the same. Javascript event is a text editor with minimal code checking. Most beginners use GDevelop because they can’t, or don’t want to, code. They want to use interfaces and change settings (like behaviours configuration) without having an understanding of what goes on behind the scenes.
gdevelop isnt just for beginners, its a game engine it should have alot of features, + this is an easy one since you can already make shaders using the js event
Sure, it would be nice and probably not too hard to implement (the editor itself), but as a backend developer of 20 years, I’m not sure I’d dare to write my own custom shaders. I can fetch data from the database tho
For starters, I’d just like to have existing (or similar) effects for 3D as we have in 2D…
One of the downsides I’ve experienced and witnessed with GDevelop effects or shaders is that they are CPU hogs, and cause noticeable pauses or slowdown in game play.
That’s with the effects that come standard with the engine. You would expect those to be as efficient and optimised as possible. Give dev with minimal shader knowledge the opportunity to create their own shaders, and many will create absolute behemoths and not understand why.
I’m not saying don’t do it, but be prepared for complaints on how laggy, glitchy, resource consuming and slow shaders are on GDevelop compared to game engine X.
It happens all the time when you give inexperienced users something powerful - they are inefficient with it. It happens at work - non-coding users writing programs that create datasets gobbling up a few 100 Gigabytes of disk space and grinding servers down to a halt.
So, you’re saying that g-develop effects are extremely laggy and they should be first optimized and then the developers could think of implementing a shader editor after they optimized the built-in effects?
