About F1 Cube
One weekend I decided I was going to learn the Rubik’s cube. Bought a cheap one, opened YouTube, watched three tutorials, and ended up stuck on my living-room floor for two hours — the cube in my hand just refused to match what was on screen.
Learning the cube turned out harder than I expected. But the hard part wasn’t the cube.
A tedious learning loop
Tutorial sites slip an ad between step 2 and step 3, so re-reading means re-scrolling past it. YouTube means constant scrubbing — the instructor’s talking pace and your hand pace never line up. The few sites that offer “enter your colours” make you click through 54 dropdowns to pick from six colours.
The worst part is when the tutorial hits an edge case. You finish a step. The teacher’s next photo is X. Yours is Y. The page has no branch for that. Start over.
I realised I wasn’t resisting the cube. I was resisting the toolchain around it.
What F1 Cube does
It fixes those four things:
- Just scan it. Open the camera, point at six faces, all 54 stickers identified automatically. No 54-click colour entry.
- Every facelet computed. The solution is derived from your actual state — there’s no “but the teacher’s didn’t look like this” branch.
- 3D mirrors every move. Each step plays out as an animation; rewind and replay at your own pace, no scrubbing.
- Two strategies. Fast Solve if you just want to see it solved (~20 moves). Layer-by-Layer (LBL) if you want to learn it yourself — seven phases, one memorable algorithm per phase.
Ads are still there — the site can’t run without them — but they live at the bottom and don’t interrupt the flow.
Who I am
Chen Jian-Lun (Nick), software engineer and co-founder of UMarket Co., Ltd., based in Taiwan. A few years writing code now, mostly application development, serverless, and browser automation. F1 Cube is a weekend project that grew out of one of those frustrating afternoons.