Guide-to-Modern-OpenGL-Functions VS moderngl

it shouldnt be, it uses opengl 3 which isnt deprecated (though after learnopengl.com, you should look at this https://github.com/fendevel/Guide-to-Modern-OpenGL-Functions to learn the more modern way of using opengl) though bugs in opengl are usually kindof difficult to track down, one time i was certain everything was right but it was only a black screen, so i deleted all of the code and rewrote it exactly the same and it worked (100% just a silly mistake on my part though) as for the root_directory.h thing, without any context i cant really help you there

https://github.com/fendevel/Guide-to-Modern-OpenGL-Functions probably has all you need to cover DSA features, but by itself you won't understand it - it basically just tells you how these modern functions replace older functions.

I personally think there are enough 3.3 tutorials out there, and that new ones should teach the modern paradigms introduced in OpenGL 4.X (DSA, immutable storage, sampler objects, vertex format, etc.). Basically, things mentioned in this post and this repo.

but other than that, if you just target desktop computers, there is no reason to go lower than 4.6 or 4.5 - if some GPU doesn't support it, it's probably too old for intensive graphics anyway. Stick with 4.5+ as DSA is worth it.

This guide teaches some of the modern OpenGL functions (particularly DSA and VertexAttribFormat). https://github.com/fendevel/Guide-to-Modern-OpenGL-Functions.

an alternative way would be DSA. Have a look here: https://github.com/fendevel/Guide-to-Modern-OpenGL-Functions and example code here https://github.com/Fr3nchK1ss/OpenGL-4.6-Hello-Triangle

Maybe that helps a bit. DSA is pretty straightforward. https://github.com/fendevel/Guide-to-Modern-OpenGL-Functions . There is a lot more to modern OpenGL than just DSA. If you want to deep dive into modern rendering (and modern OpenGL) I can suggest https://www.amazon.de/Graphics-Rendering-Cookbook-comprehensive-algorithms/dp/1838986197/ref=mp_s_a_1_1?crid=NLRU5UJZT0IH&keywords=modern+rendering+cookbook&qid=1649074763&sprefix=modern+rendering+cookbook%2Caps%2C163&sr=8-1

and then https://github.com/fendevel/Guide-to-Modern-OpenGL-Functions for an actually sane and nice way of working with OGL.

You can see how to do it here: https://github.com/moderngl/moderngl/blob/master/examples/fragment_output.py

making an executable is an absolute pain but PySide6, DearPyGui, moderngl and much more exists

Whenever I try to run examples/loading_obj_files.py from moderngl/moderngl: Modern OpenGL binding for python (github.com) , I get this error:

Not sure what rgl supports, but there's https://github.com/moderngl/moderngl - provides Pythonic wrapper around OpenGL functions.

I'm guessing minimal overhead. The README says that it is a "simplification of a subset of ModernGL with some extras that were not possible to include in ModernGL." The documentation for ModernGL (lazy-link) talks about speed only in comparison with PyOpenGL, and says that it's faster since C++ calls are "bundled together," which I assume means that ModernGL (compared to PyOpenGL) does a little more than just wrapping - for good and bad.

I would not use Pyglet, I would use ModernOpenGL. https://github.com/moderngl/moderngl It's still OpenGL but with a more pythonic interface and way less useless boilerplate. Plus you get access to modern features by default (in fact, are required, and btw modern = more than a decade old), and I believe pyglet still relies on old immediate mode opengl.

As you can see, the new version is dramatically lower than the previous one, but it is consistent with the package's tags. The downside is that this won't trigger your AUR helper to update the package, so you'll have to do it manually. You may also want to just reinstall manim as a whole to account for the new dependency version.

I am currently trying to use a compute shader to compute 2D projection of 3D points (to create a little 3D renderer in python). I am using the moderngl library, and this is my code. So my compute shader is using 2 buffers: 1 for input and 1 for output. The input is composed of 3 vec3 (to represent the 3 3D points for a triangle), and the output will be 3 2D points. The code I showed you here is for testing, which just takes the input values and put them in the output buffer. Here is my problem: