Feature Request: Basic vector math

[KaiSD]

It would be great to have the ability to add and subtract vectors without having to do it manually. It would both make the code more readable and improve the performance.
Right now adding one vector to another requires a 11-16 blocks construction (depending on type), that look complex and generate complex code, while doing the thing that in gdscript would require one simple line.

Same goes for vector-x-number multiplications.

[KaiSD]

Found out that they are pretty easy to add and added them myself where i needed tham. But I think that basic vector operations should be available without creating them manually.
But I see that the problem is slightly more complex - having all operations for all different types as separate blocks will look overwhealming in current UI, some form of convenient type selection for math blocks is needed.

[manuq]

@KaiSD yes currently the way to add or substract a Vector2 or a Vector3 is by doing it manually. This is tedious because you need to dissasemble and reassemble the vectors to add or substract its floating point components. But it has its advantages. We have considered overloading the operation blocks at one point. The problem is that if we do, we'll have to deal with types and with error handling.

For example the divide block (A) / (B), considering just the float operators and the Vector2 operators:

• (A: float) / (B: float) = C: float # Divides two floats.
• (A: float) / (B: Vector2) = Error!
• (A: Vector2) / (B: float) = C: Vector2 # Divides each component of the Vector2 by the given float.
• (A: Vector2) / (B: Vector2) = C: Vector2 # Divides each component of the Vector2 by the components of the given Vector2.

So in this case the type of the block (A) / (B) can be float or Vector2 depending on the blocks attached to the A and B slots.

Suppose you have A and B set to Vector2 so this (A) / (B) block is of type Vector2 as well, and you have it attached to another block that expects a Vector2 like "change position by (value: Vector2)". Then you change A and B to float blocks, so the block value becomes float. Then "change position by (value: Vector2)" has a float inside, which shouldn't be possible.

This is hard to explain but I hope you get the idea.

Also, if the generated code is unreadable or is bad in performance, that is not a goal of this project. See https://github.com/endlessm/godot-block-coding/blob/main/docs/PEDAGOGY.md#code-generation-is-not-a-goal

[DoomTas3r]

The blocks could be separated so that given an initial type, only the other input variant can be changed at a time:

• (float) / (float)
• (Vector2) - (Vector2)
• (Vector2) / (float or Vector2)
• (Vector3) / (float or Vector3)

I tried this using the type selector I created, but unfortunately it disables the text input making multiply (Vector2) by (float) unusable.

I'm thinking that the individual input nodes could be decoupled from the parameter input and then created and parented on instantiation or option selection. The text input would appear on top of the option input, while the option input hides when a block is snapped. Another issue is that "-" gets parsed as a number.