Journal of an Inform "Clone"

[zicklag]

Quoting from post in other thread:

Arcturus - a new programming language & compiler for the Z-machine

That’s because ive been there before. I made the same mistake back in 2018. At the time i was working on a new system. I made a game for introcomp called “Mothership”. The game worked, but was only part 1 of the story. Here are some pictures. [mo2] [mo4] My system at the time was a homoiconic language (called KL) with a very small set of primitives. Ideal, i thought, for bootstrapping an IF system. But there was a problem; it got messy fast. Here’s a small excerpt of the source code to Moth…

---

@jkj_yuio

My system at the time was a homoiconic language (called KL) with a very small set of primitives. Ideal, i thought, for bootstrapping an IF system.

@jkj_yuio

The Strand IF system would therefore consist of a complementary blend of two languages; a DSL, called Strand and a system language called KL.

Fun, thanks for sharing!

This is very similar to what I’m trying to do with Reform. I really liked the idea of a language with very few primitives.

I think maybe the most interesting part of Reform so far is the way the core rule language with, very few primitives, allows you to easily design the custom, high-level language on top. I think lisp, closure, and/or small-talk might have had similar properties, but I haven’t used them before. And my high-level language is represented strictly with “sentences”, i.e. space separated lists of string arguments, which is ideal for inform-like parsing.

So far there are two concerns:

1. The rule system will be hard to build a sufficient “standard library” for the interactive fiction world with. I think this might not actually be a problem, it might just be a lot of work. It’s intimidating to see how smart inform tries to be with the way it recognizes what the user says, counting things like scope and number, etc. I’ve been playing Impossible Bottle and it’s scaring me, but I think maybe that’s a little more complicated than most games need to be.
2. The performance might be a concern. For interactive fiction uses, I think it’s probably going to be fine. There’s not many operations the engine needs to do and I have some optimization ideas lined up, but still won’t know for sure until I get there.