Open knitting file formats are messy — and a small team is trying to clean them up

The project and the problem

A group working on the Material Programming Project is building malleable knitting software for the Kniterate, a semi‑industrial knitting machine. They met this week to turn big ambitions into concrete, technical goals: machine‑agnostic workflows, usable tooling for students, and clearer backend structures anyone can tinker with. Confusion is the emotional center here — file formats are proliferating faster than a dropped stitch, and that friction risks smothering experimentation before it begins.

The file‑format tangle

At issue are several plaintext and binary formats: kcode (.kc) — the Kniterate’s native script; knitout (.k) — an open interchange format from CMU intended to be human‑readable; Designaknit’s txt exports (often called dak); and Shima Seiki’s DAT (.dat). It has been reported that many conversions are one‑way: knitout can be converted to kcode, but the reverse isn’t currently possible. Open knitting file formats are still in a larval stage and aren’t widely used outside academia. Why care? Because, as with open‑source 3D printing and embroidery tooling, machine‑agnostic and interchangeable formats catalyze ecosystems — and that’s exactly what CAD and rapid‑prototyping communities have shown (think mods, PEMbroider, p5.embroider).

A small workaround in the Kniterate editor

The team reached out to Gerard at Kniterate about importing knitout or kcode directly. It has been reported that direct import isn’t supported, but there’s a surprise: the Kniterate editor does include a plaintext import for Designaknit .txt files via a “Load txt file pattern” layer. Those DAK txt files are oddly pictorial — they encode shaped stitch maps, carrier info, stitch types and metadata — strange but usable. It’s a pragmatic sidestep rather than a clean solution, but it’s something students can actually use in a workshop setting.

What comes next

The group’s aim is straightforward: make the backend and tooling tame enough for students, and in doing so seed a toolchain others will adopt. If they succeed, the payoff isn’t just academic; it could nudge knitting hardware toward the same kind of open, interoperable stack that transformed 3D printing. Next up in the series: a deeper dive into Kcode. Knit one, purl one — but this time with formats that play nice.

Sources: agnescameron.info, Hacker News