Make Skeleton3D bone simulator an internal child

[AThousandShips]

The documentation lists this as internal, and it makes sense to keep this internal

• Fixes: PhysicalBoneSimulator3D child being added to Skeleton3D by Engine always #95238

[akien-mga]

Thanks!

[AThousandShips]

Thank you!

[MajorMcDoom]

@AThousandShips
Hi there! I noticed that the auto-added child is still marked as active in the inspector when you inspect it in the Remote scene tree. Not sure if that actually means anything, but just thought I'd raise it.

[AThousandShips]

That didn't happen before this? I don't see how that would be affected by this change

[MajorMcDoom]

That didn't happen before this? I don't see how that would be affected by this change

This screenshot is from 4.3rc3. I don't think it's caused by this change, but I wasn't sure if this was a bug or something else, and this was the only place I thought to ask since the code setting the sim's active to false was right above the line changed by this PR and I thought you would be familiar with the code. Sorry about that. Should I open up a new issue?

[AThousandShips]

I'd say to open an issue yes, I don't know the general animation area well beyond what I've worked on, and what code I can understand, my knowledge in this case was with internal/non-internal nodes

[TokageItLab]

The internal Simulator active should be linked to animate_physical_bones. If it is not linked, it may be a bug. But if not so, no problem. See also #93504

[MajorMcDoom]

The internal Simulator active should be linked to animate_physical_bones. If it is not linked, it may be a bug. But if not so, no problem. See also #93504

I see, thanks for the clarification! I think my concern now is just that there's an additional internal modifier node that is processing just because of the default value of this "linked active state" being true, despite there being no physical bones present.

If that is not an issue though, I'll just ignore it.

[TokageItLab]

The existence of an active modifier will cause the deferred process to be called. So strictly speaking, there a very very small performance loss, so if you are really concerned about it, you should use a build with disable_deprecated = true to remove the internal modifier.

However, as I fixed in #93502, this should not be a big performance problem, since the updating does not actually occur in deferred process unless the physical bone is not in the child.