Crosscut

In programming language design, there are these islands of purity. Lisp, Forth, Smalltalk, probably others too. Those are examples of language families that implement a single concept without much compromise, resulting in a simplicity that feels desirable. Early on in my journey of working on Crosscut, I hoped that I would create something comparably simple.

And it seems like that didn't happen. I just proposed infix operators, for crying out loud. But I'm not sad. It's not like I tried and failed, after all. Along the way, trade-offs presented themselves to me. And I chose the options that steered the language away from purity, and the simplicity that comes with it.

So I chose not to create a Forth or a Lisp. (I never considered to create a Smalltalk, although Crosscut takes a lot of inspiration from it.) And I guess I rediscovered why most languages are a mess: To achieve purity, that needs to be a priority. Which means you're going to make trade-offs that sacrifice other desirable properties. Again and again, I wasn't willing to do that.