Welcome to nicescholia Discussions!

[WolfgangFahl]

We use Reel Driven Development here.

What is Reel Driven Development (RDD)?

Reel Driven Development is a software engineering method for the Semantification approach:

• A screen recording of a graph walk is the specification. The walk starts at a node of a View/Bounded Context — with cloud systems the norm, a web page. Every user action (following a link, switching a tab) visits a neighbour node, exactly the individual exploration that DDD and Specification by Example call for.
• The audio track carries the user's narration while walking: "I like this… the feature that is missing is… unfortunately here I cannot…". That links each statement directly to KANO-style requirements gathering.
• Agentic AI closes the loop: audio transcription and screenshot/frame analysis turn the reel into a modeled graph walk (hops with node, url, time, summary) plus findings — which become issues, which drive implementation, which is verified by the next reel.

Why here?

nicescholia is the RDD proof of concept for migrating a legacy system — Scholia — to a modern environment by applying named parameterized queries (#9) and the Y-Principle properly.

Findings from a reel fall into three groups, each with its own issue tracker:

1. scholia — the legacy system's own gaps (WDscholia/scholia)
2. snapquery — the named-parameterized-queries middleware (WolfgangFahl/snapquery)
3. nicescholia — the new environment (issues here)

Each reel gets its own discussion in Show and tell — starting with #12, the ScholiaAuthorWalk-Q80 example — and its artifacts live in the repo under rdd/<WalkName>/: the original video (web-readable mp4), the original transcript, the hop screenshots, and an RDD.md index linking to where the results are.