Tutorial 3 challenge - big difference between using dedicated position variable and reusing clip_position

[Pieresqi]

Hello there,
I am currently going through your Learn Wgpu and learning wgpu.

Doing 3rd challenge and comparing it with your solution I have noticed a big difference between using dedicated position variable and reusing clip_position. Reusing clip_position will give me 1 colored, light yellow, triangle but using dedicated variable will yield multi-colored triangle.

I have changed this

learn-wgpu/code/beginner/tutorial3-pipeline/src/challenge.wgsl

Line 24 in 64de216

return vec4<f32>(in.position, 0.5, 1.0);

to this:
return vec4<f32>(in.clip_position.xy, 0.5, 1.0); (or this: return vec4<f32>(in.clip_position[0], in.clip_position[1], 0.5, 1.0);, etc)

What's the cause of this big change ?

[sotrh]

There is a step in between the vertex shader and the fragment shader that converts the per-vertex data to per-pixel data. The @builtin(position) flag corresponds to the final clip position in the vertex shader, but in the fragment shader it is the pixel coordinates of the current fragment.