You don't need a note taking system.

I have started and abandoned four note taking systems. The apps were fine, every one of them. I was the one who kept quitting, and it took me a while to notice I had been quitting at the same point each time.

There is a small graveyard on my phone. A folder tree I built one Sunday in January and opened twice afterwards. A tagging scheme that ran to about 40 tags, of which I used 6, one being "misc". A weekly template that survived three weeks, which I remember because I was quite pleased with myself in week 2.

None of those apps were bad. That is the part that nagged at me for a long time. They did what they said they would, they were built by people who plainly cared about the problem, and I stopped using all of them anyway.

I don't think a system is what anybody is short of. Whatever goes wrong is happening earlier than that, in the second and a half between having a thought and doing anything at all about it.

Nobody sorts the shopping at the till

Picture the weekly shop. It goes through the checkout, it goes in bags, you get it home, and at some point that evening you put it away. Milk in the fridge, tins wherever tins live in your house, and the thing you bought for no reason sits on the counter for a week looking at you.

Nobody does that bit at the till. Nobody unpacks at the checkout with a queue forming behind them, working out whether tinned tomatoes belong under cooking or under store cupboard while the person behind shifts their weight and studies the ceiling. Bagging and putting away are two different jobs, done in two different places at two different speeds, and keeping them apart is so obvious that nobody has ever had to be taught it.

Then you open a notes app.

Which notebook, and which folder inside it? Which tags, and while we are here, is this a Project, an Area, a Resource or an Archive? Notion wants to know which database. Obsidian would like a link or two and ideally a daily note to hang it from. Every one of them is asking a reasonable question and, this is the part that never gets said out loud, asking it at the worst available moment.

You are at the till with the queue behind you and roughly 8 seconds before the thought is gone.

The label you want is written later

Filing only works if you already know why you will want the thing back. Standing at the till, you don't.

I once saved a note about a supplier putting their renewal price up. I filed it under work, which was correct and, as it turned out, completely useless. Six weeks later I wanted it back, and I wanted it because it was the thing that changed my mind about how we price a year. There is no folder called things that changed my mind. If there were one I would not have known to use it in October, because in October it was just a supplier being annoying.

The label you actually want gets written by the future, and the future is not standing in the queue with you.

Which is why the automatic tagging that has appeared in half the notes apps lands nowhere useful. It answers a question nobody has asked yet, quickly, and being confident about it is somehow the worse half. The tag you would have wanted is not knowable at the moment of typing, not by you and not by a model reading over your shoulder.

Keep capture stupid. One ugly sentence, no title, no folder, walk away.

A capture box on its own is a faster way to lose things

The trouble with what I have just written is that plenty of people say it and stop there, and stopping there does not work either. I know that, having done it for the best part of a year.

An inbox with nothing pulling things back out of it is not an inbox. It is the bags left standing in the hall, still packed, that you have quietly started stepping over on your way past. Four hundred fragments in you stop opening it, and you stop for a specific reason, which is that opening it has begun to feel like opening a bill.

The shopping does have to be put away. That was never the bit I was arguing with. The argument is only about where it happens.

Five minutes, and only once

Sunday, kettle on, I scroll back through whatever has landed in the week. Most of it I read and delete on the spot. Some of it becomes a task with a date attached. Two or three things are worth keeping properly, and by Sunday I can tell which, because I am stood on the other side of the week that made them matter.

The question I put to anything still hanging about is what does this change. If the answer is nothing, it goes, and it goes without any sense of throwing away something valuable, because a note I have read twice and never acted on has, whether I like it or not, already told me what it is. Anything untouched for a month goes the same way. If it mattered it comes back on its own, and mostly it doesn't.

Look at the shape of that for a second. Capture happens 20 times a day and costs seconds. Sorting happens once and costs 5 minutes. Almost every productivity system runs it the other way about, a small decision at every single capture and then a large reorganisation once a quarter when the thing has silted up, and the reorganisation is, every time, the point at which I quit.

The test I use now

How long does it take you to go from thought to saved? Lock screen to stored, count it honestly. More than a few seconds and you will not reach for it when it matters, and when it matters is always, without fail, when your hands are full and somebody is talking to you.

Then the second one. Six weeks on, can you find a thing back without remembering what you called it? If getting to it depends on you recalling a filing decision you made in a hurry six weeks ago, the filing was done for the app's benefit rather than yours. Search that reads the words you actually wrote is the honest version of this. A folder tree is you doing a search engine's job in advance, by hand, badly.

Where Catchlight sits

Catchlight is what taking all of the above seriously looks like, which is to say it is mostly an app with things missing from it.

There are no folders and no notebooks. Nothing to file into means no decision at the till. A Take is one thing that can be a note, a task and a reminder at once and you never pick which before you start typing, so the only thing you are doing at capture is capturing. It lands in one timeline in the order you had it. You can start one from a widget, from Control Centre, from the Action button, from Siri, who asks for the words and saves them without ever opening an editor, or from the share sheet out of whatever app you happen to be in.

Search reads what is inside a Take, so finding something back needs a word that was in the thought rather than a word you stuck on it afterwards. For the Sunday pass there are filters for notes, for tasks, for reminders and for the ones you flagged as Important, which is the bit I use most.

It is not out yet. I am building it in the open and everything above is what has been built and how it behaves, not a promise about later.

There is more on the shape of it in one Take instead of three apps, and your phone is a casino covers the other half of why the answer here keeps being to take things out rather than add them. The features page lists what is actually built, the glossary explains the words, and if you want to line the notes apps up against each other the comparison pages do that job.

The shopping still has to go away. It just doesn't have to go away at the till, with a queue behind you and the ice cream going soft in your hand.