Since my last post, I've been tormented by some mounting problems with this code. It's all decent enough, but I keep coming up with more ideas - stuff I'd like to add. Every time I come up with a new feature, the complication of the code seems to multiply. My brain keeps screaming at me that I'm doing something wrong - there's clearly something off in the way I'm implementing this design pattern. On the topic of design patterns, I'm not entirely sure which this falls into. I'm fairly new to the details of design patterns, but this must fall somewhere in the MVC/Mediator/Observer model. I suspect it's mostly observer, and that's the cause of the complications I foresee, but I'm not entirely certain.
Anyway, the problem I see is this: Games are complex. RPGs are particularly complicated, and that is my ultimate goal. To continue on my current path, it is almost impossible not to expect about 1 million events. ONE MILLION almost IDENTICAL classes. That sounds like bad design if I've ever heard of it. I've wracked my brain all week, and finally feel like I'm starting to stumble my way over this hurdle. Like so many things, there's tons of information on basic design patterns, but much much less (and far less useful) information on implementing the patterns in nontrivial examples.
So my next few posts will build up into a new direction. I'll probably stop tagging the posts with MVC, since I'm not at all sure of its accuracy. Don't worry though, I'll keep up the running list so we can follow this path without hurting ourselves. We'll now be going on a wild adventure whos outcome is uncertain, starting with some simple things like buttons, and then throwing together some simple game logic and views.
Labels: design patterns, MVC, pygame, python, series 1
Leave one.