The problem with writing your own animation software is that you have to maintain it.

The software I’m currently using is actually my second generation of animation software. The first generation was written in college for The Emperor’s New Clothes, which was done on a student research grant (let this be a lesson for you kids – you can get money for your films if you call it research!). On that project I was the writer, director, producer, and lead software engineer, so thank god Marshall did the hard animation or the whole thing would have been trash. The quality of code was pretty good for a college student (read: pretty terrible overall), but I learned some good lessons:

1. Debugging on set sucks
2. C++ is like concrete in that the code seems completely malleable when you first write it and slowly settles until the mere idea of refactoring base classes is exhausting.
3. With C++ you spend 10% of your time designing your GUI and 90% finding out why clicking a button is de-referencing a bad memory pointer.
4. If you put too much of your business logic in the GUI you will never be able to reuse it.

Back when I was in college with big dreams of film making the one book I read over and over was Robert Rodriguez’s “Rebel Without a Crew“, where he documents the making of his first film “El Mariachi.” Originally he planned to make three low budget films for the Mexican home video market, and by the time he finished the third he would be a real filmmaker. This flawless plan was undone by the small problem that El Mariachi became an independent hit and launched him into stardom.

While I wish he had done two more for Mexican home video before becoming a film wunderkind as none of his following work has had the character of El Mariachi – it’s like his “American Graffiti” – I’m starting to think that I should try his original plan. Animation is pretty difficult and it is hard to put all your eggs into one long work, especially when halfway through you get tired of the characters.

My thought is to take the same set of characters and write a five minute, fifteen minute, and forty minute scripts for them. First, the effort put into writing the five and fifteen minute scripts will be invaluable in working on the forty five minute scripts, especially when it comes to plotting and characters. Second, much of the cost of character development and production design will be front-loaded onto the first two productions, meaning most of the problems will have been solved for the forty five minute production. And if I’m bored with the characters after the first project, it will leave a nice exit point.

In other news, I went back to the KU library to do research on the films. After my initial shock of how everything that has changed at Watson since my time at KU – they have concessions now! – I browsed the dusty stacks until found the books I needed. Special thanks and one up bonus to Brian who came with me and let me use his library card to check out my research materials . The down side is that I need to rewrite everything I’ve written thus far, but I’m getting new ideas about how history can play into the story. Exciting stuff.

When I’m working on a computer program, I generally will mull a problem over for a long time, sometimes days, before a single line of code is written. This of course appears to the untrained eye that I am “juggling in my office” or “reading The Onion“, but I want to have a full understanding of the problem and understand all the implications before I begin. When I actually actually sit down at the keyboard, the initial coding often comes out in three to four hours of bezerk code rage, followed by two to three days of debugging.

In some ways the same applies to my writing. When I write stories, I generally have a fuzzy idea of what where I am going to go before one word gets written. I generally have an idea of who the people are and signposts the have to hit by the act breaks. The difference is that during the writing process I have a lot more points where I have to take a break and make a decision that affects everything going forward.

A good example is picking a name for the a character. The character’s name should be just a token that identifies them in the fantasy universe, but actually is much more. Once they have a name they have heritage, background, a family history, something kids made fun of growing up, and so much more. Making that one decision can give your character so much life they may no longer want to be bound to the fate you have planned for them.

In my case, I wanted my main character to name his dog “McGuffin”, which is something of a writers joke. I thought that I could make it work if he named it after a relative, and this introduced a Gaelic heritage to the character. Given he lives in the Old West it applies new rules to him that weren’t there before. It gives whole new areas to explore in the piece that I hadn’t planned on.

I think I will keep hacking away at my first draft, but I’ll need to go to the library as well to do some research. Three pages into the first draft and I’m actually starting to feel like a writer.