Caterpillar

The previous prototype was designed to answer specific questions: whether keeping the language extremely simple, and augmenting it with a debugger from the start, is a good way to approach Caterpillar's development. And as I teased yesterday, the answers to those questions aren't as clear to me as I had hoped.

Yes, building a debugger was practical. No, the debugger alone wasn't enough to tame the language in its super-simple form. But then, implementing the additional language features required to do that, wasn't too much of a burden. Maybe it was unreasonable to limit the language so much in the first place.

It's not clear to me, if the debugger was necessary. It took a lot of work to get it to its current state. And it will take more work to keep improving it, and to maintain its current functionality as the language evolves. I could have invested that work into language features instead, and maybe that would have gotten me a better result.