I am coming from c++ and have been using Opengl and glfw. I wanted to try out zig and build a small engine in it I started by using zig bindings(like zig_glfw and ziglgen) but i saw that zig could also interop with the original c libraries. Is doing this better than using zig bindings or should i just stick with the bindings?
It is really just a matter of taste, there isn’t really going to be any performance considerations or practical reason beyond your opinion on what “feels” better to you.
Using hand-crafted bindings can often offer more significant improvements to ergonomics, but generally this is not going to be the case with generated ones like one would expect with OpenGL. Either way, the impact from a development perspective is still relatively low.
The main benefit of using a generator is the crud of wrangling function imports across different platforms dependent on the OpenGL implementation.
When interfacing to C directly it’s easy into falling into C permeating the rest of your design, bringing along C’s flaws and footguns. Whereas when using a binding you stay in Zig.