The fact is that your GPU displays triangles, not quads. And it's almost trivial to build a rectangle of two triangles, so the API really does not need to be burdened with the possibility of rendering squares. OpenGL goes through a major cropping process, reducing a ton of functionality that made sense 15 years ago, but no longer matches how the GPU works, or how the GPU will ever work. I believe that the fixed-function pipeline has also disappeared from the latest versions, because, again, it is no longer needed, and it no longer matches the way the GPU (programmable shaders) works.
The fact is that the smaller and more complex the OpenGL API, the easier it is for sellers to write reliable, high-performance drivers, and the easier it is to learn how to correctly and efficiently use the API.
A few years ago, almost everything that was in OpenGL could be done in three different ways, which would put a lot of work on the developer to figure out which implementation is right if you want optimal performance.
Therefore, they are trying to optimize the API.
jalf Jul 10 2018-11-22T00: 00Z
source share