You don’t need to buy a system to keep your photographs in order. You’re already standing inside one — your folders. The only question is who can read it once the app is gone.
_DSC4521.ARW. The camera assigns it from a counter that means nothing to a human — and actively turns hostile over time. Reformat the card and the counter resets to zero, so five years from now you can own twenty different files all called _DSC0001, scattered across your drives, each insisting it’s the first.
And right now that file exists in exactly one place: a small piece of consumer flash memory in a slot. The most temporary storage in the entire chain. The first act of photography-after-the-shutter isn’t editing. It’s rescue — and the question of where the file lands is the most consequential decision in the whole workflow.
Getting the file off the card needs no special software — drag it in your file manager, use Apple’s Image Capture, any importer. The tool genuinely doesn’t matter. What matters is what almost everyone gets wrong by default: they let the software decide where the file lives.
So the file disappears into some app’s opaque “library,” and the photographer never forms a mental model of where their own pictures actually are. The alternative is one decision, made once: the file lands in a folder you designed, on disk, that means something with no app running. Everything else here is a consequence of which branch you take.
Rate a photo, add a keyword, set an exposure — none of it is written to your photograph. It goes into a central proprietary database. The books on the shelf are blank; one card catalog in the middle of the room is the only thing that can find them.
Now the structure lives on the things themselves — in the names, the folder tree, and a small text file beside each photo. You can walk the shelves and read. No librarian required.
The catalog’s index goes dark and its books become unfindable — you kept the pixels, lost the structure. The folder doesn’t flinch: every book still announces itself. Same failure, opposite outcome.
Once you stop letting an app decide, the shape is almost always the same — and it’s the right one, because it matches how we already keep things: one box per stretch of time, a few words on the end so you know what’s inside without opening it.
Shallow, chronological, self-labeling, obvious to a stranger — including the stranger you’ll be in twenty years. Not the only sane scheme; just a sane one. They all share that spine.
Undo the camera’s nonsense the moment the files arrive, driven by the one thing it recorded honestly: the timestamp, stamped in the EXIF at the instant the shutter fired. The date leads, always, so files sort chronologically wherever they end up; the capture counter keeps shot order within a day; the shoot rides on the end so the file announces itself anywhere.
This is profoundly un-clever, and that’s the strength. At ingest you don’t yet know which photos are worth categorising — but you know when, and when is enough to give every file a permanent, collision-proof, legible identity. (Shoot both formats? The pair shares a base name, differing only by extension — or one orphaned JPEG will drift loose in a folder of eight hundred and no tool will reunite it.)
Names and folders encode when. But a real catalog also holds ratings, keywords, captions, copyright, and the develop recipe. Where does that live without a database? Beside the file — in a small, human-readable text file (an open standard, invented with some irony by Adobe) that shares the photo’s name.
The original RAW is never modified; it stays a pristine negative. Everything you decide about it accumulates in the text alongside. This is the piece that turns a folder from a pile of files into a real database — the metadata travels with the photograph, printed on its spine. Delete the app and the text stays. Move the folder and every rating follows. There’s no central index to burn down, because the index is distributed across the files themselves.
With files named and self-describing in folders you own, a photograph’s life becomes five stages, and the discipline that makes them work is separation: at any moment you’re doing exactly one kind of thing. You cull without the temptation to edit; you develop without a library panel nagging at the edge of the screen.
Cull is the one most people skip — and the hardest rule is at its end: when you’re done, delete the rejects. Not “move them to a maybe folder.” If you don’t, you haven’t made a decision; you’ve deferred it forever and bloated the archive. And export is disposable: the permanent assets are the untouched RAW and its sidecar — everything else regenerates from them. None of these stages needs a specific app. (DigTek’s own suite splits them this way — a culler, a browser, a developer — but that’s one implementation of the principle, not the principle.)
If a culler, a library app and a developer all write to the same sidecar, why doesn’t one wipe another’s work? Because they observe a contract. The text is divided into clearly bounded namespaces — separate, labeled rooms in one document.
A well-behaved app writes only in the rooms it owns and leaves every other room untouched, byte for byte — including rooms it doesn’t even recognise. An app that honours this can join your workflow without permission and leave it intact when it goes. One that rewrites the whole file disqualifies itself, however good its features. Respect for the file is the price of admission.
“But I have a hundred thousand photos — my computer can’t read a hundred thousand text files every time I search.” It doesn’t. A folder-aware library app still builds a fast index in memory. The difference is its status.
In a catalog app the database is the source of truth — the information exists nowhere else, so if it dies, your work dies. Here that same index is merely a cache, built from the sidecars. Corrupt it and you delete it, point the app at your top folder, and it rebuilds every rating, keyword and edit from the files in minutes. The stakes of any single app’s failure collapse to nearly nothing.
Every photo application reduces to one of these postures. The whole argument is a preference among them.
The database owns the truth; your files are dumb pixels it points to. Lose the app and you lose the structure. It seizes your archive.
The software decides for you — analysis that changes what it does behind the scenes. Powerful, but a moving target you can’t fully predict or own. It outsources your judgment.
It reads and honours your folder and sidecars, does what it says, and leaves untouched what it doesn’t own. A guest that visits and leaves the place as it found it. It respects both.
Not that catalogs are evil or AI useless — both have real uses, and high-volume professionals especially may still need a catalog’s heavy machinery. It’s that for a great many photographers the third posture is available, sufficient, and quietly liberating — and it’s been hiding in plain sight, in the folders they already have. Obey, don’t outsmart. Respect, don’t seize. Own, don’t rent.
So: the folder is the database. Not a metaphor, not a downgrade, not a workaround for people who can’t afford “real” software. A thoughtfully named folder tree, with a plain-text sidecar beside every untouched negative, is a genuine, complete, searchable database — with the one property no catalog can offer: you can read it with your own eyes, on any machine, forever, and no application owns it. The catalog never made your photos more organised. It made your organisation more dependent.
None of this means working without software, and it shouldn’t. A hundred thousand photographs want an index. Reach for one, gladly — just let it be an index, not an authority: a cache it could throw away and rebuild tomorrow, never the place your work actually lives. Pointed that way, a library app is pure gain — the speed of a catalog without the captivity.
A photograph needs two things to exist: a development, and a home. The companion piece teaches the light. This one keeps the negative — somewhere you can always find it, in a database you can open with your own eyes, indexed by whatever tool you please, and owned by no one but you.
Build the home once. Apps will come and go. The folder will still be there, still readable, still yours.