Why is our shader code so much more strict?

[Pryme8]

Why is our shader code differ from like shadertoy for example where I can do something like this:

vec3 col = 0.5 + 0.5*cos(iTime+uv.xyx+vec3(0,2.0,4));

for example and have it compile just fine (notice the integers mixed with floats?)

Why does that work in a lot of other gl setups but ours freaks out?

[Evgeni_Popov]

@Pryme8

vec3 col = 0.5 + 0.5*cos(iTime+uv.xyx+vec3(0,2.0,4));

This line does work with Babylonjs:

Babylon.js Playground

See line #20.

[Pryme8]

hmmm, ok then I guess I don’t understand when you can mix floats and ints? Do you have any insight to that?

[Evgeni_Popov]

No idea! I don’t know GLSL well enough.

[Raggar]

You are not in fact mixing floats and ints, you are creating a vector3 of floats, so:

vec3 col = 0.5 + 0.5*cos(iTime+uv.xyx+vec3(0,2.0,4));

Becomes:

vec3 col = 0.5 + 0.5*cos(iTime+uv.xyx+vec3(0.0,2.0,4.0));

If you were to, say do the following:

vec3 col = 1 + 0.5*cos(iTime+uv.xyx+vec3(0,2.0,4));

You would get errors due to adding an int to a vector of floats.

[Pryme8]

So when passing ti to vec3 it inherently converted it to a float?
I guess then if it was a iVec3 it would cast to ints?