A big problem I have when developing games is thinking too much about systems, and not implementing them.  To that end, I spent several hours a few days ago implementing some sort of complex weapon material system, and then realized too late that is wasn’t necessary at all.  I also implemented some sort of special DamageModel class to control elemental weakness, which was, again, removed shortly after.  When I’m working I tend to start wondering if I am really implementing a system how it should be done, and then I bog down and have to take a break, during which I realize that the most important thing about the system is that it actually works.  It doesn’t matter if I break proper model view controller order for my monsters, or don’t have all my classes that run AI or combat as attachable subtypes of some sort of component hierarchy.  As long as it functions I’m doing well, or at least better than a quarter of what Ubisoft releases (let’s hope I don’t get sued for that).  So after a week’s work what do I have to show for it?  Well, a couple of simple, uncompleted weapon datatypes and not much else.  Maybe Ubisoft does win this round.