Reduce maintainance cost of particle nodes

[QbieShay]

Describe the project you are working on

Godot engine

Describe the problem or limitation you are having in your project

Godot has 4 ways of doing particles:

1. GPUParticles 3D
2. GPUParticles 2D
3. CPUParticles 3D
4. CPUParticles 2D

Every time I make an improvement to particle nodes I have to

1. implement it on process material
2. test it on gpuparticles3D
3. fix it fror gpuparticles3D
4. iterate
5. test on gpuparticles2D
6. fix it for gpuparticles2D
7. iterate
8. reimplement it for CPUParticles3D
9. test on CPUParticles3D
10. add conversion code from gpu to cpu particles
11. fix it
12. reimplement for CPUPartilces2D
13. add conversion code to CPUParticles2D
14. test it
15. Maybe then merge the PR
    And if changed are made with the rendering infrastructure
    Also implement it for vulkan and gles3. ( see Particle internal refactor and additions for more artistic control godot#79527 )

This cost me A TON of energy that I don't have, and it adds a 2x bug surface on Particles.

I think for all the ways Godot is easy to pick up and contribute to, particles go in the opposite direction.

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

I can see multiple ways to go about this

1. Implement Add support for particle recording and playback #7085, and make cpuparticles playback only.

2. Use a glsl to C++ converter
   It's important that it goes this way and not the other way so that people more versed in glsl than c++ can contribute as well. I think perhaps there's a way to make CPUParticles execute a process material on CPU and some automatic step to convert it.

3. Remove them since the renderer update in 4.x makes them a bit redundant (Deprecate CPUParticles #7517)

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

I would prefer option 2 of the option specified above, even if it means we have to do our own code generation. particles are something very important for artists and it being one of the most laborious nodes to maintain, considering it's an artist feature, makes it incredibly hard to get new contributors on, leaving higher bus factor and maintainance cost.

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?

Seems like particles cannot be addons due to the way they need deep engine core interaction.

[webredraptors]

2. Use a glsl to C++ converter
   It's important that it goes this way and not the other way so that people more versed in glsl than c++ can contribute as
   well. I think perhaps there's a way to make CPUParticles execute a process material on CPU and some automatic step to convert it.

Could you elaborate on how you think the CPU could execute ParticleProcessMaterial? I was under the impression that PPMs were essentially just shaders that the user can program without using shader code. Implementing and maintaining a full GLSL to C++ converter sounds like a monumental task.

[QbieShay]

Implementing and maintaining a full GLSL to C++ converter sounds like a monumental task.

This is what I'm suggesting, and it doesn't really need to be a full blown glsl to c++, it just needs to be the subset that gdshaders use in particle process. However i do not know effectively how big of a task that is.