The Curse of Closed Formats

Here at Kioko, we fight customer lock-in. We are fully committed to open design files and interoperability.

A problem worth solving

Every closed format started as an honest fix for a real problem.

In the early 1990s, moving a document between two computers was a gamble. Fonts got substituted, margins drifted, a figure on page three landed halfway off page four. PDF fixed it: a container that froze a document exactly as written, pixel for pixel, on any machine. It worked, and the world adopted it.

CAD had the same problem one dimension up.

• Same geometry everywhere — A part designed in one system had to mean the same surfaces, same tolerances, same intent — or a multi-million-pound assembly wouldn't fit together. CATIA, NX, SolidWorks all solved it, storing not just a shape but a full parametric feature history, exact to the micron. For decades they were the only serious option.
• Size mattered — A complex part is a dense mess of surfaces, edges, and control points, and when disk and memory were scarce, packing all of it into a compact binary blob that parsed fast was good engineering. Binary won on the numbers.

These formats existed because the problems were real — exactness and efficiency.

Where the curse sets in

Sadly, a proprietary binary format is a one-way door, and three costs follow — each compounding the last:

• Lock-in. Your entire design history sits in a format only one company can read natively. Switching tools means abandoning or painfully converting decades of work. The cost of leaving is the product.
• The interoperability tax. Since the native formats are closed, the industry trades geometry through neutral intermediaries — STEP, IGES, Parasolid. Every translation loses information: parametric features, design history, and intent get stripped down to dumb geometry.
• Archival rot. A part designed in 2003 may not open cleanly today. The format is undocumented, the old licence is dead, and the geometry is a fossil — present but unreadable. For industries that certify parts for thirty-year service lives, that's not a nuisance, it's a liability.

None of this serves the engineer. It serves the incumbent. A closed format is a monopoly mechanism: it makes your data the hostage and the software the ransom.

Why PDF escaped and CAD did not

Here's the parallel. PDF was just as proprietary as any CAD format — and then it wasn't.

In 2008 Adobe handed PDF to the ISO. It became ISO 32000, an open standard anyone could implement. Adobe didn't die; an ecosystem exploded. Browsers render PDF natively; phones, libraries, printers, and a thousand tools read and write it. The format outgrew its owner and became infrastructure — and Adobe kept thriving by building better tools on top of an open base.

CAD never had its 2008. The native formats are still closed, undocumented, and tied to the licence, and the neutral formats that fill the gap are lossy by design. Engineers get to choose between fidelity inside a walled garden and freedom in a lossy desert.

That's why the community is moving away. Not for fashion, but because engineers have seen what an open format does for documents and know the same can be achieved for geometry.

The text-to-CAD solution

We need an alternative that is open, legible, owned by the engineer and can withstand the test of time.

At Kioko, openness is a first principle, not a feature granted later. A CAD model should be something you own outright: small enough to store anywhere, readable without a licence, versionable like code, and legible thirty years from now without an archaeology project.

And here's the part that surprises people. The old size argument for binary was real, but it answered the wrong question.

The legacy formats are large because they store the evaluated result — every surface and edge, fully baked. Kioko stores the recipe instead: scripts that regenerate the geometry on demand. So the text-based model isn't a tax you pay for openness — it's the more efficient representation. A part that's tens of megabytes as a baked binary is a few kilobytes of script: small enough to email, to diff, to track in Git next to your code. Legibility and compactness at once — where the binary formats traded one away and never delivered the other.

The lesson of PDF is simple. A format becomes infrastructure the moment it stops being a hostage. Documents learned this two decades ago. Geometry is overdue.

If you want to help build CAD that engineers actually own, write to us at contact@kioko.io.