gemmapeter.co.uk

This week has seen some interesting code. I’m currently working on my MEDA302 (Graphics) coursework mostly this has meant me brushing up on my C++ skills and figuring out how to use the SDL and SDL_gfx libraries. The coursework is really interesting, to make a simple 3D program that does scaling, translation, rotation and a volume sweep. The catch is that I’m not allowed to use a 3D API, the polygons have to be drawn on the screen using my code to project the 3D object into 2D screen co-ordinates.

I was worrying about optimising the memory usage but then I thought “Why?”, my program only displays a cube at the moment (using 12 triangles), modern games use millions of triangles and they still run at a resonable speed. Although they use specialised hardware and my program puts the calculation code through the CPU instead. I don’t think memory optimisation will make that much difference with the number of triangles I’m using. My program is just a “Hello World” compared to modern 3D games.