The whole concept of versioning software that is built using GenAI is that things change fast and in large ways. In the before times, large feature additions would be considered, designed, and took a long time to deliver. With the models progression from Opus 4 in December of '25

Over the last two weeks I have aligned Sharpee with a new parser-IF modeling language called Chord. This effort was not to adhere to one side or the other. It was a true alignment where both sides were required to maintain their "way", be elegant, and in

After the first pass at designing Chord, I followed the same step as I did with Sharpee. I worked with Claude to build first an Author's Guide, then documentation for the Standard Library. The first document used Family Zoo as its spine and although we found a few

Now that version 2.2 of Sharpee is shipped with the website and developer manual updated, I took a swing at developing a more author friendly syntax that compiles to Sharpee. I'm calling it Chord. It's currently implemented in a branch and working with Cloak of

Phrase Algebra Fully Implemented * Sharpee 2.x replaces the entire text-rendering system. Text is no longer a string with embedded formatters; it's a tree of typed Phrase values that defer realization, rendered by one per-locale Assembler at the end of the turn. * The old formatter chain

There is still one glaring weakness in Sharpee directly related to separating logic from text emission. Sharpee is intentionally a developer-first/author-second platform and I've leaned into that design. I often use the words, "The Sharpee Way," when talking to Claude. As an exercise

If you hate GenAI, you probably want to skip this episode, though I would assume anyone reading my blog should already know what I do. Sharpee has been stable-ish for a few months. I bumped it to 1.0 and started revising the repository documentation. It occurred to me

Update: Even though the packages are working for most aspects of Sharpee, I am going through an alternate regression process and finding more issues. I'm already bumping it to 1.1.0 and likely to hit 1.2 this week. I restructured the Family Zoo tutorial to act

Even though the platform has been stable for a couple of months, it has shown gaps as I work on stories and extended use cases. One of those use cases was inspired by an intfiction.org thread which asked about multi-user play where two or more people could play