Why does Color::rgb use f32 instead of u8?

[wyhaya]

u8:

Color::rgb(0, 0, 0);
Color::rgb(255, 255, 255);
Color::rgb(3, 169, 244);

In many cases, rgb uses numbers from 0-255, and using u8 will not appear invalid colors

f32:

Color::rgb(312312321.312, 312312.0, 654645.1231);

f32 cannot verify the correctness of the value

[karroffel]

What values are useful depends on the situation. HDR rendering and HDR images for example (high dynamic range) need to be able to go above 1.0, so going with f32 means you can have higher dynamic range than when using a normalised u8.

[wyhaya]

@karroffel Thank you for your reply.

This is a very common one. Is it necessary to add Color:from_u8(255, 255, 255) or similar methods? It should be more convenient to use rgb in this way.

[karroffel]

I'm sure a function like that can be added, I wouldn't think that should be a problem :)

I personally don't use normalised u8s often to describe colour values, to me a range from 0.0 to 1.0 is more intuitive, unless you are dealing with serialisation code or direct image formats. But if a from_u8 methods makes code clearer in some cases then I would expect that to be a welcome addition

[cart]

Yeah we do need floats in some cases. But from_u8 would be very welcome!

[ncallaway]

I'd suggest normalizing with the Color::rgb and Color::rgba functions, so something like Color::rgb_u8() and Color::rgba_u8()