In October 2025 I wrote that the habit matters more than the software. I still believe the first half of that sentence, and the second half turned into Scratchpad.
The piece was Documenting the Mess, on my own blog, and the advice was simple. Write while it still hurts. Keep the failed attempts, the angry cancellation emails, the ugly screenshots, and the 2am ideas.
I told everyone to use whatever they already had open. For me that was Apple Notes.
A year later I've written every single day.
Most of it went out on that blog and on DailyTips.Dev, where I write about building products and software. My two photography newsletters, DailyPhotoTips and TheDailyPreset, took the rest, and anything left over sits in a pile of private notes nobody will ever read.
I've also spent most of that year building a notes app, amongst other things.
That's a strange thing to do for someone who really believes the software doesn't matter, so I wanted to work out what actually happened.
A Confession
Looking back, Documenting the Mess was mostly a confession.
When I stop writing, things go wrong in a very specific way.
When I stop writing, things go wrong in a very specific way.
A decision I made in March gets made again in September, the opposite way, with total confidence both times.
And an idea I was sure I'd remember is just gone. All I'm left with is the feeling that there used to be something there.
Writing is the only thing I've found that stops this from happening.
Some people need a calendar for the same reason. Without one the days still happen, they just happen to you.
Forgetting and Carrying
When I sat down and asked myself why I need this, the first answer was obvious. I forget. Everyone does, and faster than they think.
Hermann Ebbinghaus measured it on himself in the 1880s, memorizing lists of nonsense syllables and testing how much survived. Most of it was gone within a day. Whatever I learned last Tuesday about why a deploy failed won't be there in six months, no matter how much it hurt at the time.
My memory edits, too. After something works out, my brain quietly rewrites the story into one where I knew all along. The note I wrote during the mess is the only witness that wasn't in on it.
The second answer took longer to see. I carry things.
An unfinished thought keeps tapping you on the shoulder, usually in the shower or at 1am.
In 2011 the psychologists E.J. Masicampo and Roy Baumeister found that making a concrete plan for an unfinished goal was enough to quiet those nagging thoughts, even before the goal was done. Writing it down is that plan. Once it's on the page, I'm allowed to put it down.
So I'd describe writing as how I hold on to things without having to hold them in my head.
The Pace Changed
A year ago I was building at human speed. I'd decide something and sleep on it before building it the next week. The thinking and the doing were far enough apart that some of it stuck on its own.
That's not how I work anymore.
I wrote about this in Judgment Bottleneck. AI makes me faster most days, and my own judgment is now the slow part.
What I didn't say in that piece is what speed does to memory.
When a feature goes from idea to production in an afternoon, nothing has time to settle. I make a dozen decisions before dinner, each one reasonable in the moment, and by Friday I can't tell you why half of them went the way they did.
Git keeps every change I made, each with a one-line message that rarely says why.
The faster I build, the more I need to write.
My notes are the only place those reasons survive.
I've already written about how writing is where my thinking actually happens, in Unwritten Ideas Don't Count. This is the other half of that idea. Even the thinking I get right evaporates if it doesn't land somewhere.
The Software Mattered More Than I Said
Here's the part I got wrong.
"The habit matters more than the software" is true on day one. If you're choosing between writing in a mediocre app tonight and researching the perfect app for three weeks, write tonight. I'd still say that to anyone.
After a few months of notes, finding things becomes the hard part.
After a few months you have hundreds of notes, and finding things becomes the hard part.
The notes that matter most are usually the ones you wrote in a hurry, with half a sentence and a PDF someone sent you.
A plain notes app is worst at giving those back.
What I needed turned out to be a short list.
- Search that reads everything. The quote I want is usually in the PDF, and the error is usually in the screenshot. If search only reads what I typed, it misses most of what I saved.
- Every version kept. A note that changes over months is a record of how my thinking changed. Overwriting it destroys the very thing Documenting the Mess was about.
- Voice that turns into text. Half my best ideas show up when my hands are busy. A memo I can't search is a memo I'll never play back.
- One calm place. I wanted somewhere fast to put a thought before it leaves, with nothing to set up first.
There's nothing exotic on that list.
I just couldn't find all of it together in something that felt calm, so I built it.
Designing for One, Again
I wrote about how this goes in Designing for One.
Scratchpad started as a tool for exactly one person. It ran on my own server and did things in the order I needed them.
Then my family wanted in, and after them a few friends.
Before long I was writing empty states and help pages for strangers, which is exactly the job that piece warns you about.
Building for one feels like skipping the hard design work. You're really just taking a loan against it, and the empathy comes due with interest.
The difference this time is that I knew the second job was coming, and I wanted it. A notes app that only works for the person who built it isn't much use to anyone else.
This week I finally moved my own old Apple Notes across, with every picture and recording still inside them.
That closes the circle for me. The app I recommended a year ago is now the app I'm importing from.
What Goes in My Notes Now
The original list still holds. I keep writing down technical decisions, failed attempts, harsh feedback, and the numbers that made me change direction.
A year of doing it added a few.
- I write down the reason. "Switched the queue to SQLite" is useless in six months. "Switched the queue to SQLite because Redis was one more thing to babysit on a box I already babysit" is a decision I can revisit.
- I keep what I almost did. The option I rejected is often more useful later than the one I picked, because it's the one I'll be tempted by again.
- I note what I told people. Promises to customers and prices I quoted come back eventually, and they come back with a date attached.
- I add how it felt. One line is enough, about shipping something or watching it fail. Six months later that line explains more than the changelog does.
Most of it is badly written, and I've stopped caring.
Somewhere to Land
The headline on the Scratchpad homepage says your next great idea needs somewhere to land. I wrote that line about the app. It turns out it's also the most honest answer I have to why I write.
Ideas don't wait for you to be ready. Mine show up mid-deploy, or on a walk with the dog.
If there's nowhere to put them, they just pass through.
If there's nowhere to put them, they just pass through, and you're left with the vague sense that you had a good one once.
Writing is how I catch them, and how I find them again later.
A year ago I said start tonight, and I'd still say it. Open whatever you've got and write down the thing that's bothering you about the project you're working on right now.
If you keep at it, at some point you'll care a lot more about where all of it lives.
I did.
