The first public alpha introduced Permutrix as a tool born from a real production problem: combining portrait parts, colors, positions, and transforms without rebuilding every result by hand.
From game portraits to a standalone tool
Permutrix began as portrait-generation logic inside a game project. Once the system could combine image parts and vary color, position, and transforms, it became clear that the same structure could help with webtoon assets, game art, character variations, and other repeatable visual work.
That small internal system gradually became a separate layer-combination tool built around reusable artwork and explicit creative rules.
The difficult parts of the first alpha
- Joint calculations had to carry parent movement into child layers without losing the newly shifted pivot.
- Color variation needed enough range to feel different while preserving the original artwork's lightness and character.
- Asset organization had to remain understandable as the number of available parts grew.
- Batch generation and large pattern layouts needed to scale beyond a handful of results.
A drawing sewing machine
The long-term metaphor in the first announcement was a drawing sewing machine: a tool that lets creators weave their own image parts into larger systems instead of placing every piece manually.
That idea still describes the center of Permutrix. The artist draws the material and chooses the rules; the software helps reveal, compare, and export the variations inside that system.
Looking back from 1.2.1
The first alpha was intentionally rough, but it established the product identity that remains today: user-owned assets, controlled variation, reproducible structure, and no generative-AI artwork.