CGI during my spare time
I’ve been doing trying the tutorials from The Nature of Code by Daniel Shiffman.
I fell in love with Computer-Generated Graphics (CGI) during my masters in Digital Media at Georgia Tech. In fact, I first learned to program in Processing. However, outside of the programming courses I teach at Michigan, I’ve not had much time to actually dive into CGI after I finished my masters. It’s been eight years and I wanted to hone my skills. So, I took advantage of my social distancing time productively.
I first started to adapt scripts from the processing tutorials into pygame on my own. However, I quickly realized this was not getting me far. For once, I was just translating code from Java (processing is after all Java). I felt like I was translating a math paper from English to Spanish. Sure, I could make every sentence syntactically and grammatically correct, but this was not helping me grasp the argument, or in this case, the logic underpinning the commands. Thus, I decided to go back to the origins. I learned that Shiffman had an online version of his book and grabbed onto it.
It probably would be much faster if I just followed the instructions with Processing, but that’s too easy. More importantly, Processing is explicitly designed to abstract away the heavy lifting of setting up the environment for CGI. Pygame does the same to an extent, but much lesser. If I want, for instance, to instantiate a Euclidean vector object in python, there is no built-in PVector class. I have to search for vector classes, and sometimes still extend them with additional functionality of my own. In so doing, I’m making sure I master the ins and outs of python syntax structure, and getting a healthy reminder about basic trigonometry and physics. These learning outcomes will help me in the future, if I want to write a CGI script in a different language.
In the mid-term, my hope is to use my learning to extend available options for information visualization through other python libraries like numpy. If I get away with this, I can actually link my current experiments with my plans to combine data-driven storytelling and my dissertation research on social memory. The way is long, but I’m not in a rush. After all, the priority right now is my dissertation. The good thing about CGI is that I find it so absorbing that it completely takes my mind out of my research-related concerns at the end of the day. So, I deepen my skills as a programmer, learn new ways to generate graphics, decompress, and make cool graphics. Thanks, Daniel Shiffman!
If you want to check out the scripts I’m producing, stop by my repo here. I would love feedback to make better scripts.
Here is the very first script of this experiment, the classic bouncing ball: