There's a version of the story where it all starts clean. Where someone brilliant draws something perfect on the first try, and the engine just works.

That's not how it happened.

The real version starts with a list. Handwritten, half-legible, in red ink on a sheet that's already been folded twice and stained with something that might be oil or might be coffee. Lean surge flagged between five and a half and six and a half thousand RPM. Cylinder four running hot — outer fin stack, reduced airflow, needs its own AFR trim. TPS alignment off by 1.2 millimetres. Clutch dragging on cold start. Coating batch questioned. Intake stacks showing peak efficiency at twelve-two, but the valvetrain can't hold that speed yet.

Not a blueprint. A confession.

Every rider who has ever cracked open a service manual at midnight, trying to figure out why the idle won't settle or why the fuelling feels off above seven thousand, knows this document. Not this specific one. But this kind. The kind that says: something is wrong, and I'm writing it down so the next version of me — or the next person — doesn't have to find it the hard way.

The inline-4 motorcycle engine didn't arrive. It was argued into existence. Four cylinders mounted sideways across a frame that was never designed for something that wide. Engineers shaving millimetres off bore diameter just to fit the thing between the rider's knees. Going undersquare — longer stroke, narrower bore — because a wider engine simply didn't clear. Accepting the combustion trade-off at high RPM because the alternative was an engine that didn't fit the motorcycle.

And then, once it fit, discovering that four separate carburettors feeding four separate cylinders don't synchronise themselves. That the outer cylinders run hotter than the inner ones. That rocker arms wear at the tip and the valve stem cap, and eventually, at sustained speed, a valve drops.

None of this was failure. All of it was information.

The test log is not what engineers write when something goes wrong. It's what they write when they're paying close enough attention to notice. Lean surge at 5,500 RPM doesn't mean the engine is broken. It means someone rode it through that band, felt the hesitation, measured it, flagged it, and handed it to the next shift with a note that says: fix this.

That's what development actually looks like. Not inspiration. Annotation.

The engines that came after — the ones that scream past fourteen thousand RPM and make two hundred horsepower and rev so clean you forget there are explosions happening inside — they exist because someone sat in a workshop with a crankshaft that wasn't ready, wrote down every way it was wrong, and kept going.

The TEST LOG from UBF's IL4 collection carries that document. A crankshaft assembly. Red ink annotations. Not a celebration of what the engine became. A record of what it was before anyone understood it.

The most important version of the engine is always the one that doesn't work yet.