The Hobbit Hole

Notation is imporant. After having dabbled and learned a dozen programming languages, this one thing has become certain to me.

Not only in computer languages though, but in languages of every type. How you write affects what you write. There is the theory that one’s language constrains how one thinks, but the Sapir-Whorf theory is no longer considered valid. Still, it stands to reason there is at least some truth therein, especially when considering the aesthetics of a language. Few argue that French is a beautiful language, and even fewer claim that it is ugly.

That much said, why isn’t there a notation for game play? Well, there wasn’t, until now. Danc at the Lost Garden has started exactly that. It has more in common with music notation than conventional written language, but it is indeed an impressive start. It also highlights that thinking about gameplay is a completely different endeavor.

I haven’t digested it completely, but it certainly looks promising. I hope to be able to use it while writing my own game.

It seems that Lisp has gained a bit of traction lately, perhaps in part due to Practical Common Lisp being nominated for a Jolt Award. As well, reddit.com, being initially written in Lisp, has garnered it some attention, and Paul Graham, with his essays on it, is perhaps more responsible than anyone for Lisp’s (small) resurgence.

It has also gotten to me again too. I’m a sucker for a good language. I myself cannot attest to the productivity gains due to the language, but I do find something alluring in its simplicity. Still, I’ve decided to at least experiment in tying my engine to the Lisp language, and even putting some of the lower level features in Lisp first. I’ve already done some prototyping in it, and it proves versatile and fast enough for now.

I’m particuarly intrigued by GOAL, the Lisp dialect used by Naughty Dog. True, they no longer use it, but I would love to get a look at how it worked. Scott Shumaker, a lead programmer at Naughty Dog, gave a brief overview in this message. I find it interesting that, according to him, the code generated wasn’t that optimized. Perhaps GOAL made assembly that much easier to write, negating the need for the optimizing? Sounds like it to me from reading his commentary.

This does remind me of a project I was rather proud of at Singletrac. When writing the PSX renderer we used in some of the later games, I wrote a little sexp parser that had primitives that equated to one instruction each. I would then run it and it would spit out the corresponding assembly. I still had to debug it in assembler, but it was much easier to write initially due to that representation. And man, was that renderer fast. Clocked against some of the higher profile titles of the day, it would easily match the most high performant renderers. But alas, it went into a game that sold at most 30,000 units.

So, from personal experience, I can say that some form of Lisp has helped me, though to be honest, I would hardly state that my sexp parser was a true lisp dialect any more than I would say that XML is a lisp dialect.

Oh well, here’s hoping that using Lisp will turn me into a super productive programmer and I can get this game/book project done that much sooner!

I laughed.  Hope you find it as amusing.

So here’s why I’m enjoying the fact that I studied math in college.

Business week has an article about the power and price of math. Does this mean that mathematicians will be the ones that teenagers idolize? Not likely, but now is a good time to study mathematics.

Seems a bit like psychohistory to me, though I do believe that math is a language that can describe just about any phenomenon.