This week I finally understood the difference between motion and direction. Everything was moving. Nothing was finishing.
The honest update
Zero commits on the 3D car montage project again. It has now been frozen since August 27 — two full months.
Instead, the week produced commits in three other places: a newspaper-style redesign of TeknoPulse with a read-later feature, security headers and accessible pagination; the move of BenihSaham into a pnpm workspace monorepo; and the second devlog post for this very project, published on October 21.
Four repos on the board this week — and not one of them closer to done than it was on Monday.
What actually went wrong
Last entry I wrote that I’d pick one of two options for the car sim this week: resume with the smallest real step, or pause it formally with a resume date. I did neither. I didn’t write down a decision at all.
What I did instead was keep four things half-alive at once. That felt productive, because every day had visible output: a commit, a build, a deploy. But none of that output compounded. Switching contexts every day means paying the setup cost over and over and never getting the payoff of deep work — the kind where a problem stays in your head overnight and gets solved while you brush your teeth.
The real problem isn’t which project
For months I’ve been treating this as a project-selection problem. Wrong project, wrong idea, wrong timing — keep looking. But the pattern is the same regardless of the project: I start well, get to the unglamorous middle, and then something new starts running in parallel. Three months, several repos, and the same result.
The bottleneck is not the idea. It’s that I never give a single thing enough consecutive attention for it to become good.
So the decision this week isn’t “the car sim or something else.” It’s “one thing, on purpose, for long enough to matter.” I’m picking an iOS mobile app: one app, one milestone, and the whole week’s attention on it.
What happens to the rest
Formally paused, not abandoned. The car montage project goes on the paused list with a written resume trigger instead of sitting there pretending to be active. Same for the other repos — they keep their state, they lose the pressure.
That distinction matters. Paused means it’s someone’s turn later. Abandoned means closing the door. What I’ve been doing is neither, which is how a project can occupy space in your head for two months without receiving a single hour of work.
Why this isn’t just another switch
Fair question — I’ve now got three “new focuses” in three months. Two guardrails against repeating that:
- A minimum before reconsidering. Two weeks of real work before any talk of switching, unless something external breaks and forces the issue.
- One active project, written down. Not a mental note. A line in the contract that says which project currently owns my attention, and a paused list next to it. If something new wants in, it goes to the parking lot and waits a week.
The first milestone for the iOS app is small on purpose: the app builds and runs on a physical device, with its core flow working end-to-end. Not launch, not App Store. Just proof that attention applied to one thing produces something you can hold.
Motion is easy. Direction is the hard part. This week, one direction.