Shader Coding - Best Practices
Jump to navigation
Jump to search
Work in progress This article or section will be worked on in the upcoming hours or days. Note: Will be based on: http://www.mail-archive.com/flightgear-devel@lists.sourceforge.net/msg35934.html See history for the latest developments. |
Supported GLSL Versions
Note Be aware that we currently _DO_NOT_ support glsl > 1.2x. Please don't mix '''glsl''' flavours: please stick to glsl 1.2x.(even though support for it is largely deprecated). The transition to OpenGL 3.0 / glsl >=1.3x is a very big undertaking, and not backward compatible. It is always a good idea to declare the expected GLSL version at the top of each file:
#version 120
|
Vertex Shaders
Note For testing purposes, you can use a simple pass-through vertex shader:
#version 120
void main(void) {
gl_Position = ftransform();
}
|
Fragment Shaders
Note To check if your shader is working, add this as the last line, it should turn all pixels black:
gl_FragColor = vec4 (0.0,0.0,0.0,1.0);
|
Note if4dnf: it's usually bad practice to do any kind of operations on gl_FragColor assignment. At most a vec4() swizzle is accepted, although even that one is treated differently based on the platform (some assign a temporary variable, some don't). Do wahtever you need in a separate vec4 variable and just assign it's value to gl_FragColor at the end. |
Note If the fragment isn't running, there should be an error message in the log file (I wish they'd still be written to the console). Might be that the texture isn't defined in the supporting framework for instance - the shader assumed that it is available, but you need to declare it in the matching C++ code/effect file first. |
Note This is how you can make a pixel darker:
gl_FragColor = color * 0.5;
|