thank you. That’s what I think I did, but for some reasons I don’t get what I’m expecting, even though the metal gpu trace suggest that some results are ok, but not displayed correctly, it’s very strange.
The buffers are the same across pipelines… I will test it, but I don’t need to rebind buffers after having bound a new pipeline, do I? In other words, the bindings are not supposed to be order-dependent?
(I guess that like a gazillions time before in my experience of graphics programming, my code has a silly error that once fixed, makes everything render correctly :-/)
The buffers are the same across pipelines… I will test it, but I don’t need to rebind buffers after having bound a new pipeline, do I? In other words, the bindings are not supposed to be order-dependent?
no they aren’t…
(I guess that like a gazillions time before in my experience of graphics programming, my code has a silly error that once fixed, makes everything render correctly :-/)
…turned out the culprit was not that silly, but now it works as expected, thanks again for your help!
This is a picture I took of testgl.c running. Plain old OpenGL.
It looks worse when you take a picture because the camera isn’t synced to your monitor, so the result has multiple frames from your monitor combined into one photo.
But I don’t necessarily need a camera, because the effect is visible to the naked eye.
I don’t see anything with OpenGL and with SDL_gpu like in my first photo.