PoseEstimationForMobile VS Cocos2d

Tonal has a pretty beefy android system in it. https://github.com/edvardHua/PoseEstimationForMobile

That's all I can really tell you from what you've posted. The original version without the Blowfish encryption is open source (eg. https://github.com/cocos2d/cocos2d-x/blob/v4/cocos/scripting/lua-bindings/manual/CCLuaStack.cpp) , and I don't really understand how why you have so much symbol information available if what you're trying to disassemble isn't also probably open source

As others have pointed out, the biggest difference you're seeing is likely due to the _version_ of OpenGL (and hence GLSL) version. That said, there are still important differences. I'd recommend looking at a comparison between the same shader in a project that supports both OpenGL & OpenGLES. For example, here's a shader from cocos2d-x https://github.com/cocos2d/cocos2d-x/blob/v4/cocos/renderer/shaders/positionColor.vert. Note the only difference in this case is the additional precision qualifier (lowp) for v_fragmentColor. Note too how cocos uses preprocessor macros to handle this, so they don't have to maintain separate shader sources. Depending on your goals, you might be interested in tools like Nvidia cg or nvFX that allow for creating shaders in a dialect agnostic way, but ymmv.

Speaking about markdown files on GitHub, I know that there are some repositories that have very stunning README markdown files, such as this one for Cocos2d-x, whcih includes images in conjunction with text. I might do the same for my repository as well. Considering VS Code have an extension for .md files, this might come in handy when I consider using VS Code to contribute to my game projects.