Reverse Camera2D.rotating to ignore_rotation

[Mickeon]

As partially brought up in #54161 (comment).

rotating is misleading, as Camera2D is affected by rotation and global_rotation like any other Node2D.

This PR reverses the boolean outright, and makes it clearer the Camera2D's view ignores its own rotation when it is now enabled.

This PR has been refactored. It used to simply rename rotating to rotating_view.

[golddotasksquestions]

What does this property do anyway and how do I use it?

The description does not really help. From this description I would assume this property locks the Camera2D onto a target and makes sure the Camera2D rotation follows the target. But there is no property to set a target.

If this would be a method instead and would take a target argument, I would also better understand, but that's not the case here either.

So all in all I don't see how rotating_view is any more helpful than rotating. Both names do nothing to help me understand intrinsically and the description is not helping either.

[Mickeon]

I was considering overhauling Camera2D's descriptions (again) and names in general because they honestly feel misleading. Indeed, there's no such thing as a "camera target" because Camera2D merely inherits its transform from its parent, like any other Node2D. Renaming the property is a step in the right direction, as should be possibly renaming its other properties as well.

In this PR #55699 the explanations were heavily limited by how badly named some properties are, as well as how they've been previously described. There's no consistency now.

[KoBeWi]

You could update the description in this PR, to make it more complete.

[Mickeon]

Updated this specific description to be more explicit, too. Tell me what you may think.

[golddotasksquestions]

If true, the camera's rendered view is affected by its Node2D.rotation and Node2D.global_rotation.

Yeah that's a lot better! Thanks!

[golddotasksquestions]

On second thought, maybe also add rotation_degrees? Since this is the only one exposed in the Editor Inspector, and therefore maybe easier to find for people who want to test this property.
If [code]true[/code], the camera's rendered view is affected by its [member Node2D.rotation_degrees], [member Node2D.rotation] and [member Node2D.global_rotation_degrees], [member Node2D.global_rotation].

[Mickeon]

There is no rotation_degrees and global_rotation_degrees in Godot 4. Even if there were, one could argue that description to be needlessly verbose.

[aaronfranke]

How about can_rotate? This would make sense to me. If it's false, it can't rotate, so any rotation is discarded.

[Ansraer]

During the meeting it was agreed that this definitely needs a rename, but we couldn't agree on anything. Suggestions and discussion are welcome.

[KoBeWi]

What if instead we reverse the property and call it ignore_rotation?

[Mickeon]

That sounds great to me! It could also be called rotation_ignored so that the getter suits more. And perhaps it could all tidly fit into a "Rotation" group when #65776 is merged.

[Mickeon]

I was going to push the rename, but I realised that all other Camera2D have "enabled" booleans, instead. But at the same time, rotation_enabled feels wrong? Where to go from here?

[Mickeon]

The PR has been refactored. This PR reverses the boolean outright, and makes it clearer the Camera2D's view ignores its own rotation when it is now enabled.

[KoBeWi]

Looks ok, but I didn't test the converter code.

[Mickeon]

I am unable to test the converter code, but if given instructions on how to I could give it a try. I know right off the bat that it cannot be perfect on every situation, but fortunately, rotating is something that is very unlikely to be changed in code...

[akien-mga]

Thanks!

[Mickeon]

Reminder that this is worth accentuating with big caps someday because it's not just a simple rename this time around.