GPU Particles (2D + 3D) visibility rect / gizmo optimization for reduced visual clutter

[RPicster]

As I am dev with a very "visual and effects" heavy approach, I often use a lot of Particle effects.
This was first proposed here: godotengine/godot-proposals#3554

Problem in 2D

In the current version this often leads to very hard-to-understand scenes because of all the rendered visibility rects:

While this is an artificial example, my real-world examples are much, much worse.

Problem in 3D

While I did not do a full 3D project so far, I did one test scene in 4.0 and the gizmos of 3D Particle systems were a big annoyance to me. When I am building scenes I want a clear vision of how it will look in the game and big, colored gizmos overlaying the particle system I am working on in that moment are rather obstructive.

I found out that I can change the gizmo color in the editor settings, but it is not having any effect on the 'solid' color when selecting the gizmo.

My changes do two things

GPUParticles2D:

Only selected particles will show their visibility rect box, this reduces visual clutter on the stage.

bToQZtji4H.mp4

GPUParticles3D:

The 'solid' color takes the original gizmo color into consideration for the alpha.
Before a user could set the Particle3D gizmo color to transparent, but it would still always show the solid box upon selection because the alpha value was hardcoded to be set to 0.1.

With this change, the gizmo colors alpha from the editor setting will be taken into consideration and the alpha will be 0 when below a certain threshold.

SM52zZDS8Z.mp4

This is my first "real" PR and I'm a C++ newbie, so if there are things that could be improved, I'm happy to hear about it!

[GuyUnger]

i like this a lot, these boxes clutter up the scene without adding very valuable information when not working with them. hope it gets merged

[RPicster]

Seems I forgot to include editor_node.h in the includes of the gpu_particles_2d header file...
Just pushed the fix to make the checks hopefully work.

[KoBeWi]

The include should be in either cpp or header, not in both. And cpp is preferred if possible, because includes in header cause unnecessary recompilation of files when any dependency is modified.

[RPicster]

Ok, I notice I still have a lot to learn 🤣

The failed builds for the templates were caused by parts of the new additions not being wrapped in #ifdef TOOLS_ENABLED.
So parts were included when building the export templates.

I changed that and also tested building export templates using tools=no target=release which build without errors.

I also included the change requested by @KoBeWi

[RPicster]

I started over with my implementation as I was pointed to the problem in regard to this decision: #29730

The visibility rect visibility is now controlled by the gpu_particle_editor_plugin and the dependecy to editor is removed from gpu_particle_2d

[RPicster]

I unexposed the set_show_visibility_rect method from the scripting API, it wasn't neccessary and would create the need for documentation.

[akien-mga]

Looks mostly good, see comments. The commits should also be squashed together, see PR workflow.

[RPicster]

Looks mostly good, see comments. The commits should also be squashed together, see PR workflow.

Alright! All commits are squashed and the review feedback was done!

[KoBeWi]

Looks ok, aside from 1 more nitpick.

[RPicster]

Looks ok, aside from 1 more nitpick.

I'm thankful for all of your feedback! It's very new for me to write c++ code so I use every opportunity to learn!
I pushed the change :)

Changes done.

[akien-mga]

Thanks! And congrats for your first merged Godot contribution 🎉

[RPicster]

Thanks! And congrats for your first merged Godot contribution 🎉

Thank you! Was a pleasure!

[RPicster]

I was looking through my merged PRs for potential 3.x backports. I would love to have this in a future 3.x version - shall I create a backport? @akien-mga @KoBeWi

[akien-mga]

Yeah that sounds good 👍