Up to this point I’ve tried to write on a more open schedule, pushing for graphical updates (since everyone loves pretty pictures).  That hasn’t quite worked out for me, since I’ve lately been working on a couple side projects and mostly deep engine stuff.  Instead, I thought I’d try a more stream of consciousness textual update style for a bit and see how that suits me.  Probably pepper that with some screenshots when I have something interesting to show for it.

So, I’m currently re-working the bullet engine to more closely match the design pattern of BulletML, basically making my own flavor of the system that works better in C# using the content pipeline for loading and working with patterns written in xml, as well as hooking into my GPU based particle system using bullets as individual emitters.  This means I’m also working on a fairly optimized version of a tiny expression parser to handle expression based values as read in from the bullet xml descriptions.  Allowing me to write “pi” in a value field instead of hard coding “3.141592″ as well as a “rand” value in expressions which will generate a new random number in the expression every time the bullet hits that part of the script.  I hope to post more about my bullet engine when it’s more fully formed, including code and descriptions of design decisions, because I think it’s fairly interesting.