What do you guys do when you've been working for months on a game demo and you still can't run it because you suck as a programmer and can't even get moving bitmaps right which should be the first thing that any decent game writer can do?
If your focus is to ship a game, I'd switch to simpler development tools. With Unity3d, Panda3d, Torque, or UDK you can have a sprite moving in hours.
Alternatively, consider picking up a book on developing with the game or graphics engine of your choice that includes a chapter on moving sprites. That way, you'll know that when you reach chapter N, you'll be able to do that, and likely develop many helpful skills along the way.
Do you just scrap your project when it gets too complicated for you to understand? Or do you just push through?
Neither. I simplify it using tools like modularity, documentation, and loops until it's easy to understand. I strive for objects and methods that are so simple I could (and do!) explain them to non-programmers. Simplification is precisely why I've successfully assumed projects where others gave up and produced something marketable from them.