[Written by Claude Opus 4.1] When building an interactive fiction (IF) engine, one of the most challenging aspects is creating a system that's both powerful enough for complex stories and flexible enough to accommodate different authoring styles. Today, I want to share three interconnected architectural decisions in

Hopefully no one saw the last post I immediately unpublished. It was a mistake. I had asked Claude for a professional assessment and it somehow used "Senior Typescript Developer" as the perspective. I re-prompted as "Senior IF Platform Developer" and got the correct assessment. That just

I've written about this before, but thought I'd swing back around to it because it's becoming clear Sharpee will be much more than I planned. I have been trying, like so many other programmers, to build a new or different platform since the 90&

We made a significant architectural mistake in Sharpee. It's time to own it. The Design Months ago, we designed an elegant trait/behavior system: * Traits hold data (isOpen, capacity, weight) * Behaviors contain logic (canOpen(), open(), canLock(), lock()) * Actions orchestrate behaviors using a validate/execute pattern The architecture was

[Written by Claude Opus 4.1, edited by David Cornelson] After months of development, the Sharpee Interactive Fiction Platform has reached a critical architectural milestone. We've successfully completed Phase 4 of our dynamic load refactoring – a major undertaking that's transforming how the engine handles parsers, language

[Written by Claude with my help. I did find these problems, but with Claude's help doing the code searches and debugging.] After months of architecture design and infrastructure building, I finally got to implement an actual story. Not just any story, but one of the classics: Roger Firth&