Chaos Trees

Have you ever done a lot of work writing code over tree based structures?  I've been doing that a lot lately. It's kind of a habit now.  It started years ago working on the .Net XML library, and it has progressed through various language development projects.  I've spent the last few weeks trying to get a particular transformation just right.  The code manages translating one tree structure representing language A into another tree structure representing language B with a variety of contortions along the way.  Its hard to think about this kind of problem because you only get to tinker with a little bit of the transformation at a time, there are so many subtleties.  Even trivial differences you make in an early part of the transformation can become unwieldy and unrecognizable at later stages.  Definitely non-linear.  It's like my own little bit of Chaos Theory.

I'll try to avoid comparison to the fabled butterfly effect, that's been abused by Hollywood.  Though, it does remind me of Stephen Wolfram's work on finite automata.  My code is definitely a finite state machine, and the result of its functioning given particular inputs will to the casual observer appear to be complex.  It seems absurd to consider that all of reality may be the outcome of some finite rules, almost as if it were pre-ordained that I'd be sitting at my computer writing these other finite rules.  Whoa, slipping into philosophy.  I can't be talking about multiple-world theory in one post and absolute fate in another.  Or can I?  I suppose each offshoot reality could have its own determinism, that fate in each branch is sealed, except for the fact that it keeps on branching due to those probability waves.  That would mean that it was fate in this reality that I'd write that code, but it was only by chance that I got here in the first place.  What about that other poor sucker who is still struggling to figure out the algorithm?  We really should figure out a way to communicate between these realities, and let them all know I've already done it.  I mean if Hollywood can come up with Sliders, then the world of science ought to be able to get it to work, right?

But then, if reality is controlled by finite automata, then there really is no uncertainty to afford a probability function, and I suppose there would been no branching in the fabric of space-time. I wonder if this is what happens to a person if you spend way too much time writing little bits of computer code.  You start to think you can see 'the Matrix.'  I wonder if Stephen had this problem doing his research.  Was he writing a lot of code over tree based structures?

But I digress

Matt