Crosscut

It's done! I went through all the problems of the previous prototypes. As promised, it took weeks. I also explained how I want to address most of them in the new prototype, except for syntax and evaluation model. Those will be the next topic. But first, I have some news. I've started another new prototype! (A bit over a week ago already.)

This latest prototype is closely based on the "new" one, and everything I've been saying about the "new" prototype also applies to the latest one. So I hope the confusion will be somewhat limited. The only difference is the aforementioned syntax and evaluation model, which I plan to start explaining tomorrow. (I have now archived the old prototype, by the way.)

I think that after such a long period of stability (working on a single prototype), this kind of churn is pretty normal. At least it was the same after I concluded the first long-running prototype. A new approach brings new ideas. And pretty quickly, some of those turn out to be undesirable, for one reason or another.