Yes, it is not that performant. I'm trying to increase the performance but i am bound to the vertex limit
Here is an updated screenshot, rendered on my NVIDIA. The previous screenshot was rendered with an Intel 4000.
I've added texture mapping support and improved block updating. Now you can change any block any time and the chunks will be updated correctly. A worst case update includes 4 chunks, in best case only one.


EDIT: I had fps_max = 60; in my main loop... It runs with more than 60 FPS i already started to wonder...