Well here I am months later. I've gone in depth with OpenGL and discovered how difficult an API it is. The required mathematics and physics (which I've never been good at) make it really hard to achieve even something logically simple as moving a camera. Not even yet on the topic of integrating networking into a 3D application, or handling IO let alone lighting.
After discovering the difficulty level of 3D programming (Direct3D is even more convoluted), I wondered if 2D would be just as difficult while allowing multi-platform support. That was the problem with 3D APIs, theres really only OpenGL for everything. We bring it up to an easier level of understanding, and there are tools such as Unreal/Unity/Source/Unigine/CryEngine/C4, but these are limited to pretty much Windows and maybe iOS (except for Unigine, and upcoming Unity 4). These also have a hefty licence for someone whos just starting out.
So, back to 2D game development. What games today would sell if it's 2D and on the Desktop platform? I needed to know if it was a worthwhile endeavour to turn in this direction. After trying a couple of games, these two 2D games gave me some hope. Violetland, which is free and opensource is really addictive yet has really simple graphics. There is night and day in the game, and multiple effects with different weapons and a leveling system to spice up gameplay. Theres definitely more room for improvement, but its amazing for a free 2D game.
Terraria. Terraria is the real game changer for me. I couldn't believe that I spent 1 week indulged in that game with 2 other friends. The game on its own is addictive but the multiplayer co-op/pvp factor really made the replay value high. Even now while I'm busy, I itch to get back to playing it with my pals.
So, I decided to find tools I could use to develop a multi platform 2D game (since Terraria is stuck on Windows using XNA, and has halted development). I've rediscovered PyGame, and after all that OpenGL, I've found it to be really easy and robust. I'm able to code simple functionality for certain things I wish to try out (such as making a GUI or testing out tiling) in a period of half an hour to 1 hour only.
Not to mention that for physics there exists PyMunk, and for networking there is Twisted. I've already started a design document and am adding more information to it daily at the same time testing/learning what can or can't be achieved in PyGame. I know nothing will materialise immediately on the horizon (especially with my Java3D, VRML and OpenGL projects this trimester), but with this document I hope to continually revise my plans over the next year all the while fleshing out my understanding of PyGame, PyMunk and hopefully Twisted as well. I hope that two years from now I have something, even if only a small demo, to show for my plans and studies thus far.
