Field Note #14 ∷ On the Use of Time
“So teach us to number our days, that we may apply our hearts unto wisdom.”
— Psalm 90:12 (KJV)
Some years ago, for reasons I no longer remember, I went back to Remember the Milk. I had used it more than a decade earlier, as the centre of my increasingly elaborate attempt to get my life under control: projects, contexts, due dates, recurring tasks, lists within lists. Eventually I had abandoned it, as I had abandoned several systems before and would abandon several after.
The account was still there. So were the tasks.
What stopped me cold was how many of them I recognised. A surprising fraction, perhaps thirty or forty per cent, were substantially the same things currently haunting whatever system I had migrated to in the intervening years. Some had changed names or acquired more sophisticated descriptions. Some had been carried through jobs and houses and different stages of life. But there they were, ten years earlier, waiting patiently.
I had built an extraordinarily reliable system for transporting unreasonable expectations through time. The software had done its job perfectly.
It remembered.
I¶
In my twenties, on the old USENIX SAGE mailing list, Tom Limoncelli urged me to read David Allen’s Getting Things Done. He would later write Time Management for System Administrators. I nodded seriously and did not read it.
In my early thirties, shortly after my daughter was born, I eventually picked up a copy in an airport bookshop. I read it straight through on a transoceanic flight, turned back to the beginning, and read it again.
When I returned to Brussels, I took two weeks off and implemented the system properly. And I mean properly. I collected everything. I processed inboxes. I defined projects. I established contexts. I built review routines. I embraced the idea that the human brain is much better at having ideas than holding them.
For many years afterward, my life was dominated by todo lists. It became a running joke at home. My first smartphone was an HTC with a little trackball. I bought it largely because I wanted to synchronise Remember the Milk between my work browser and my pocket. This seemed entirely reasonable at the time. Perhaps it was.
The underlying problem was real. Human beings carry an astonishing number of half-formed intentions: send the letter, fix the tap, renew the certificate, call your mother, investigate the strange noise, read the paper, book the dentist, finish the proposal, learn French properly this time. Writing these things down is better than repeatedly remembering them at three in the morning. GTD taught me the enormous relief of getting commitments out of my head.
What I was slower to learn was that capture solves only half the problem. Once captured, an intention acquires persistence, and computers are particularly good at persistence. A task can survive indefinitely while demanding almost nothing from the system that stores it. A few bytes in a database will happily preserve an aspiration long after the circumstances that produced it have vanished.
Humans are different. We have to keep encountering the thing. Every time I reviewed a list containing some ancient undone task, I was being asked, however faintly, to answer for it again. Why haven’t you done this? Do you still intend to? Shouldn’t you?
Over the years I developed a rhythm familiar to anyone who has spent too much time in productivity systems. About every twelve to eighteen months, I would declare bankruptcy. The old system had become cluttered, stale, morally radioactive. The new system would be cleaner. This time I would maintain it properly.
There were scraps of paper. Legal pads. Text files. Web applications. Desktop applications. Phones. Wikis. Later, increasingly sophisticated notebooks. The important tasks came with me. So did hundreds of things that had once seemed important.
That old Remember the Milk account was the archaeological layer that made the pattern visible. Many of the tasks I thought I was failing to do had already survived ten years of evidence that I was not, in fact, going to do them.
I had mistaken persistence for obligation.
II¶
In those years I discovered Bullet Journalling. I adapted it fairly heavily, used it for years, and taught versions of it to colleagues.
It had one virtue I did not appreciate immediately. Migration was expensive. If I reached the end of a notebook page and still wanted an unfinished task to follow me, I had to copy it again by hand. There was friction, and this turned out to be useful. Every few pages, the notebook quietly asked whether I was really prepared to write the same damned thing one more time.
Sometimes the answer was yes. Often it was not. Nothing exploded. The task simply stopped travelling.
This was perhaps my first useful lesson in letting go. I owe it, and the habit of a pencil in my pocket, to Ryder Carroll, who designed that friction in on purpose.
We have an enormous vocabulary for commitment: goals, deliverables, promises, milestones, objectives, resolutions, action items, priorities. We have rather less language for the moment when an intention has served its purpose and should be allowed to die.
There are tasks we fail. There are also tasks we reconsider. Those are not the same thing, and a sane system should be able to represent the difference.
Every human week contains 168 hours. I think of those hours as a gift from God. A gift we must somehow learn to steward.
They are not distributed equally in any meaningful social sense. Wealth can buy control over more of them. Poverty, illness, duty, and circumstance can take that control away. A parent caring for a sick child does not possess Tuesday in the same way as a wealthy retiree with a housekeeper and no dependants. But nobody gets to save an unused Tuesday for later. The hours pass whether we choose deliberately or not.
As I approached forty, I became increasingly conscious that time was my principal asset. Money could, at least in theory, be earned again. A reputation could be repaired. A failed project could be restarted. Tuesday was going to become Wednesday exactly once.
This made todo lists begin to look strange. A todo list can grow without bound.
A week cannot.
There is no meaningful upper limit on the number of things I can sincerely intend to do. The upper limit on the number I can actually do is severe. That mismatch matters. Every yes consumes a piece of something finite.
Sometimes the cost is obvious. If I spend Thursday afternoon in one meeting, I cannot spend the same Thursday afternoon writing. Often the displacement is quieter. Work gets the hour. The walk does not. The client gets the evening. The book does not. The urgent request gets Saturday. Someone I love gets whatever remains.
A life is shaped by its refusals at least as surely as by its ambitions. Success therefore depends partly on learning what to leave undone. This is harder than it sounds.
III¶
I had known for years what I thought the answer should be: put the work on the calendar. If something is a real commitment, then at some point it has to occur in time. “I should write the report” is an intention. “Tuesday, 14:00–16:00: write the report” has collided with physics.
I wanted to become the sort of person who did that. I repeatedly failed.
At my best, I would spend several hours meticulously laying out the next few days, perhaps an entire week. The plan looked magnificent. Then Wednesday happened. A meeting moved. Something took twice as long as expected. Somebody needed help. I was tired. I became fascinated by the wrong problem. A piece of work that looked simple turned out not to be. By Thursday, the calendar was fiction. By Friday, I had stopped looking at it.
After a few rounds of this I would conclude that time blocking did not work for me, or that I lacked the discipline required to make it work. Then, six months later, I would try again. Planning became another project haunting me.
The mistake, I think, was deeper than choosing the wrong application or method. I was building plans that could not survive contact with my life. The schedule worked only if Future Me behaved approximately as predicted. Future Me has never signed such an agreement.
Human beings underestimate tasks. We overestimate energy. We forget dependencies. We get sick. Children need things. Trains stop. People ring with terrible news and wonderful news. Sometimes we spend Tuesday doing something else.
A useful plan has to survive being wrong. This sounds obvious when stated as an engineering requirement. It took me years to apply it to myself.
One of the principles I have come to value in systems design is simple:
Make survivable mistakes.
Errors are inevitable. The interesting question is what they cost. A resilient system does not require every component to behave perfectly. It limits the consequences when one does not. Why should a system for managing a human week demand more reliability from its operator than we would demand from infrastructure? A missed focus block should cost a reschedule. It should not cost faith in the entire system.
Limits are useful information. Once human limits are treated as facts, they become design constraints. Capacity can be allocated. Margin can be protected. Failure can be contained.
IV¶
There was another clue, much older.
At my first proper job, around 2000, I discovered that I did unusually good work between roughly two and four in the afternoon. So I began blocking those hours on my calendar. The events were private. I gave them all the same title:
Quantum time manipulation research.
This was my little joke. Someone looking at my calendar could see that I was unavailable. They did not need to know why. For twenty years, in one form or another, I kept coming back to this trick.
It worked because meetings had a protocol. Focus did not.
A meeting invitation arrives with machinery behind it: a start time, an end time, a list of attendees, a conference link, reminders, recurrence rules, negotiation, acceptance, rejection. The fact that I meant to spend an afternoon thinking had none of these protections, until I represented it as a meeting. Then the infrastructure defended it.
That asymmetry has become more pronounced over time. The modern workplace is extraordinarily good at allowing other people to place claims upon our attention. The claims arrive through email, Teams, Slack, ticket queues, issue trackers, calendars, phones, and whatever new channel somebody has decided will finally reduce the number of channels. An invitation can cross the planet in milliseconds and find the exact remaining rectangle of white space in your calendar.
The things that sustain a human life are often much worse at asserting themselves. Nobody opens a priority-one incident because you have not been outside all week. There is no escalation policy for calling your brother. The guitar in the corner has terrible project management. Your dog is unlikely to submit a Jira ticket explaining that the week’s walking SLA has been unacceptable.
The systems around us are therefore capable of producing a peculiar illusion. Anything that can interrupt us appears urgent. Anything that cannot waits quietly.
Quiet things can be important for a very long time.
V¶
The COVID years made this worse.
At the Belgian national CERT, the computer emergency response team the country calls when its systems go badly wrong, almost everything seemed to matter. Sometimes it really did. The boundaries between work and home blurred. The world was frightened. People were isolated. Cybersecurity incidents did not stop because a pandemic had begun.
Time lost texture.
I gave that work far more than I had to give. I burned out while entirely confident that burnout happened to other people.
Burnout is easily mistaken for exhaustion, the sort of thing a fortnight’s holiday fixes. What I experienced was closer to an injury: a poorly understood condition with physical and cognitive effects. For me it ended in pain every time I touched a computer. Skills I had relied on for years, and thought of as strengths, were suddenly out of reach, as if the connections to them had been cut. Getting them back took the better part of five years, with doctors, therapists, and a great deal of patient self-care. No amount of effort made the healing go faster.
I had choices in how much I gave, and I made some poor ones. I could triage incidents and vulnerabilities. I was much worse at triaging claims on myself.
I mention the burnout because it changed how I understand warnings. Warnings are easiest to dismiss while the system still appears to be working. The load creeps up quietly. The margin disappears before the failure becomes visible.
I had nothing then that showed me my weeks as they really were. A calendar honest about my own capacity might at least have made that harder to ignore.
Then came a large consultancy.
After a national CERT, most work feels quieter. I know several people who have tried to leave CERT life and been drawn back by the work. What surprised me was that the urgency did not change. At the CERT, much of the work felt consequential and everything was urgent. At the consultancy, the stakes felt lower and everything was still equally urgent.
The calendar looked much the same from the inside. I could be quadruple-booked for stretches of the day, jumping from Teams call to Teams call, trying to decide in each round who I was going to disappoint this time.
There is a point at which urgency stops conveying information. If every task is priority one, priority one has become the default. If every free half-hour can be booked, free time becomes merely unclaimed capacity.
By then I knew there had to be a better way. I also knew I was unlikely to become, through sheer force of character, the perfectly disciplined person my previous systems seemed to require.
One day I remembered quantum time manipulation research. And in a small fit of inspiration I opened Claude.
“Hey buddy,” I asked, “you grok .ics files, don’t you?”
VI¶
The first version was embarrassingly simple. I wanted blocks. Lots of them. Pre-made. I wanted them to fit together across my afternoons like Tetris pieces. I wanted to be able to import them into a calendar and then take whichever ones I needed. Most importantly, I wanted tomorrow’s structure to exist before Future Me had to invent it.
Claude generated the files. I dragged one into my calendar, and an entire quarter appeared.
I still remember the feeling. There was suddenly something to choose from. This mattered more than I expected. A blank calendar asks a surprisingly difficult question: What should you do with your life next Thursday at two? A pre-existing block asks an easier one: Do you want this?
My plans had always treated Future Me as someone to be cared for, though not entirely trusted. For most of my adult life I have tried to do him a favour whenever I could. As I watched the cascade of blocks fill the coming quarter, I suspected Future Me would be grateful for this one.
Over time the hack grew. Focus blocks acquired a regular rhythm: ninety-six minutes, in my case. This is not based on a secret neuroscientific optimum. Ninety-six minutes is approximately four sides of vinyl: a double album. When I work from home, the record flips give me little natural interruptions. Stand up. Turn the record over. Sit back down. No timer required.
Meeting slots followed, each UTC half-hour with its own stable emoji, so that people can find a time by looking for free animals.
This is ridiculous. It also just works.
The hack became a pattern. Eventually it acquired a name: calmoji.
VII¶
What ended up mattering most was a boundary.
Calmoji keeps two kinds of calendar. The first is a map. It contains possibilities: open focus blocks, possible meeting windows, the broader season of the year. The map can be switched on while planning and hidden while working. The second contains commitments. Those are mine.
When I decide to use an open block, I copy it into my own calendar. The copying matters. It is the deliberate act that turns a possibility into a commitment, so availability never silently becomes obligation.
The distinction sounds almost trivial. My todo systems had never made it. They had filled themselves with possibilities and then presented the result back to me as debt.
Read this book. Learn this language. Write this paper. Fix this thing. Investigate that idea. Call this person.
Maybe. Someday. Should. Could. Perhaps.
The list did not know the difference. The calendar does. There are only so many places to put things. Eventually it asks a rude and useful question:
When?
And if there is no plausible answer, another question follows:
Are you actually going to do this?
Over time I added five little markers. 🎯 means planned. 🧵 means worked on and continuing, with the next block already claimed. ✅ means done. 🏆 means a win worth remembering. And 🕊️ means released.
The dove may be the most important one. It means:
I meant to do this. Circumstances changed. I have considered it again. I am letting it go.
This is different from pretending the task never existed. It is also different from carrying it indefinitely as evidence against myself. My old Remember the Milk account had been a museum of good intentions. The collection could only grow. Release gave it a graceful exit.
That changed the emotional character of planning more than I expected. An old 🎯 in the past contains information. Maybe I underestimated the work. Maybe I did something more important. Maybe the task was badly defined. Maybe I no longer care. The useful response is investigation.
Shame contributes surprisingly little to scheduling. (I have never found “should” a helpful word, in life or in standards.)
VIII¶
There is another rule now. The calendar must survive neglect.
Sometimes I plan carefully every week. Sometimes I do not look at the system properly for a month. Life happens. Older versions of me would have interpreted this as system failure.
The ritual now is much simpler. Come back. Search for the old targets and loose threads. Some things are already done. Some belong inside something current. A few still matter and need a new time. Most can probably be released. Then start from where you are. Do not reconstruct the missing weeks. Do not punish the present because the past was untidy.
This is graceful degradation for a human being. The system assumes its operator will occasionally fail to operate it.
Good. So do I.
A similar principle governs the week itself. Put the anchors in first: meals, movement, care for whoever or whatever depends on you, errands, travel, a marker that says work stops here. Then fit focus around the life, rather than constructing an immaculate workweek and hoping a human existence can survive in the cracks.
This was uncomfortable for me at first. Paid work comes with people asking when it will be finished. Exercise usually does not. Your children may ask for you, but often at precisely the moment when something else has acquired a deadline.
Prayer, contemplation, reading, friendship, music, maintenance, recovery: these are easy to treat as residual activities. Whatever time is left.
There is rarely much left.
IX¶
Recently I added an optional AI prompt to calmoji. This may sound slightly perverse, since one attraction of the whole system is that it needs nothing beyond an ordinary calendar.
But we live in the age we live in, and many of us now have access to machines that are exceptionally good at looking across a pile of commitments and helping us reason about them.
I called the assistant an agenda steward. The word matters. Stewardship implies care, limits, and responsibility for something the steward does not own.
The instructions begin with an explicit prohibition:
Your job is not to get more work out of me.
It is there to help make the week legible. It can notice that a plan contains eighteen hours of work and eleven available, or that the people I love are receiving only leftover time.
But it cannot claim the block. It cannot make the commitment. That remains mine.
One instruction near the end says:
If the plan looks inhuman, fix the plan rather than asking the human to become more machine-like.
I wrote that for the AI. I probably needed to hear it myself twenty years ago.
X¶
None of this produces more time. That is important.
Every human week still contains 168 hours. We can improve our control over some of them. We can waste fewer. We can protect a few from accidental capture. But there is no productivity system that repeals mortality.
There is good news in this. Mortality forces choice. Choice creates shape. Immortality and infinite possibility are obviously appealing. But a life lived without the need to choose, the need to relinquish, the need to say “no” to one thing in order to say “yes” to another would, over time, become indistinguishable from every other possible life.
We are made of choices and refusals. Not everything fits. The unchosen hour disappears just as surely as the chosen one is spent. That leaves us with the harder question: what are we choosing to carry, and what are we choosing to put down?
I have spent a substantial part of the last few years thinking about time in another sense: NTP eras, Unix timestamps, GPS week numbers, Kerberos dates, the strange and sometimes dangerous design assumptions hidden inside machines about what year it is and how far into the future an integer can count.
Much of that work matters. Some of it may matter very much. But there is something faintly comic about spending one’s days worrying whether computers will know what time it is in 2038 while repeatedly discovering at six in the evening that one has forgotten what Tuesday was for.
Infrastructure needs coherent time. So do people. The human-scale problem is harder.
A machine mostly needs to know when it is. We have to decide what the time is for.
Twenty-six years after my first job, quantum time manipulation research is still on my calendar. The joke has become slightly more elaborate. Now there are open blocks waiting underneath it. Some become writing. Some become client work. Some become reading, administration, or thinking. Some remain empty. That is allowed too.
Calmoji has not made me perfectly organised. Thank God.
I still underestimate things. I still drift. I still become absorbed in interesting problems at the expense of useful ones. I still routinely discover that I have promised more week than the week contains.
The difference is that the mistakes are cheaper now. The map survives. The commitments remain legible. Things can move. Things can end. And things that no longer belong to me can finally be released.
If you would like to try it, calmoji is free and open source: plain .ics files in a zip, covering 2026 to 2039, that import into almost any calendar.
🕊️
— Trey
If you received this via email, the canonical archive version lives here: propertools.be/fieldnotes/field-note-14-on-the-use-of-time/