L-Systems

Problem

A small set of rules can produce enormous complexity — that’s the promise of Lindenmayer systems. The engineering question is how to expose formal grammars as a practical Blender tool without making users write parser code.

Approach

Implement a string-rewriting engine inside a Blender add-on: users define an axiom and production rules, the system expands them over N iterations, and turtle interpretation converts the resulting symbol strings into geometry.

Architecture

• Grammar engine — axiom + production rules expanded iteratively (Koch curves, snowflakes, dragon curves, fractal plants)
• Turtle interpreter — symbol stream drives position/heading state to generate curves and branches
• State management — push/pop semantics for branching structures
• Geometry output — interpreted paths become Blender curve and mesh objects

Engineering Decisions

• Grammar-first design: the text format is the interface; everything downstream is deterministic interpretation.
• Iteration depth exposed to the user: complexity is a parameter, not a fixed setting.

Result

A tool where typing a handful of characters produces Koch snowflakes, dragon curves, and fractal plants — complex geometry from compact rules.

What I Learned

• Recursive systems are simple to specify and easy to get subtly wrong; state discipline matters.
• Formal grammars are one of the highest leverage-per-line-of-code tools in procedural generation.