Crosscut

I think there's a lot more to say on the evaluation model (and syntax, which is a closely connected topic). Especially how to handle more complex cases, which is going to require bindings, and combinators of some sort. (And I have lots of open questions about the latter.) But I don't want to go into it now. It's time to wrap up all this talk about the new prototypes.

And to do that, I want to take a step back. So far, I've been talking about technical details, and how those differ between prototypes. Let's leave that behind, and look at what it actually is that I'm working on, and why.

And because I don't want to make this thought too long, I'll start with that tomorrow.