I know, I'm building up to it. I've already updated it so it follows a tokenize/compile/execute pattern, and it no longer uses ridiculous pattern matching as tokens; it still uses pattern matching for compiling and interpreting, but not for executing a compiled program. Now that I have it working a bit more reasonably, I'm going to finish the default syntax config and then set up a customization api.

Hey sorry to bump this, but I just finished fixing the compiler; it's still not written very well, but I'm working on that. Compiled code runs around 75-100% faster than interpreted code (at least for the tests I've been running), and now it actually runs exactly the same as if it were interpreted. Soon I'll be adding an API to modify and add on to the language.