Catalog without custody

Somewhere in your house there is a drive you are afraid to open. Maybe it is an external disk labeled "PHOTOS — SORT LATER" in optimistic marker. Maybe it is a whole laptop. You know roughly what is on it: your life, in no particular order.

And every piece of organizing advice starts the same way: first, upload everything to our cloud.

I would like to talk you out of that.

The upload tax

Say it really is sixty gigabytes. Uploading that means days of babysitting a progress bar, an internet plan groaning under the load, and then a monthly fee — forever — for a second copy of files you already had. Worse, at the end of it, nothing has actually been organized. Your originals are still sitting exactly where they were. You have paid to duplicate the mess.

Here is the part nobody says out loud: uploading was never the point. You did not want your files to move. You wanted to know what you have.

Those are different problems. I only make you solve the second one.

Libraries figured this out a century ago

The old card catalog at the library never contained a single book. It was a cabinet of index cards, and each card said two things: here is what this book is, and here is where it lives. A librarian could answer "do we have it?" and "where is it?" without walking a single shelf. The catalog was small, fast, and searchable precisely because it did not hold the books.

I work the same way, with one upgrade the librarians would have envied: my cards cannot lie about which book they belong to.

Fingerprints, in plain English

When I meet a file, I run its contents through a well-known piece of math that produces a sixty-four-character code. I call it a fingerprint, because it behaves like one. The same file produces the same code every single time, on any computer, forever. Change one pixel, and the code changes completely. If two files anywhere in your life share a fingerprint, they are the same file — no matter what they are named or which folder they are hiding in.

So a fingerprint is a library card the book writes about itself. Once it is in my catalog, I can recognize that exact file anywhere — renamed, copied onto a third drive, buried in a folder called "New folder (7)" — without keeping the file itself.

Photos get a second, softer fingerprint on top: a visual print, taken from the picture rather than the raw bytes. It stays nearly the same when a photo is resized or re-saved, which is how I catch the "same shot, slightly smaller" copies that exact fingerprints cannot see.

What actually travels: the card, not the book

The practical version of all this is a small helper program, also called Sally, that runs on your computer. You start it with one command — sally go — and it looks in the places files actually live: your photo and document folders, cloud-sync folders like Dropbox, iCloud Drive, OneDrive, and Google Drive, and any external drives you have plugged in. It shows you what it found and asks before scanning anything. It knows to skip the junk — caches, apps, system folders, code.

Then, for each file, it does the work in place. It computes the fingerprint right there on your machine. For photos, it also reads the hidden notes the camera left inside — when it was taken, where, with what — locally, on your computer. What it sends me is the card: the fingerprint, the name and size, where the file lives, those facts, a few automatic labels, and the tiny preview picture that is already embedded inside most photos. That preview is about ten kilobytes — the size of a short email.

So for a typical four-megabyte photo, what travels is a fraction of one percent of it. The photo stays home. The card comes to the catalog. The little preview is what lets your library show real pictures instead of gray boxes, and what lets me match lookalike copies by eye. If even that feels like too much, there is a --no-thumbs switch, and a --dry-run that sends nothing at all and just tells you what would happen.

What a cabinet full of cards buys you

Quite a lot, it turns out.

You get one searchable library across everything — the laptop, the drive in the drawer, the cloud folders — even though the files never left their homes. You can finally see that the drive holds 14,000 photos, that 3,000 of them also live in Dropbox, and that the same fingerprint showing up in three places means one photo with three addresses, not three photos. Dates and places read from the photos themselves line everything up on a timeline. And when I ask you my easy questions — what should we call this, when was it, roughly — you answer from the previews, without touching the originals.

If one day you do want me to hold onto the actual bytes — say the old drive starts making that noise — you can upload the files that matter, and I will recognize each one instantly by its fingerprint and upgrade its card in place. It works in the other direction too: when you drag files into the web uploader, your browser fingerprints them first, so bytes I already know are never uploaded twice.

Who this is for

The person with the shoebox drive and no idea what is on it. The person on rural internet, or a data cap, for whom "just upload it" is a two-week sentence. The person who does not love the idea of handing their entire life to someone else's server just to get an inventory of it. The person with a terabyte of video, where uploading is simply never going to happen.

And honestly: anyone who has not decided what to keep yet. You cannot make keep-or-toss decisions about files you cannot see. Custody can come later, for the things that earn it. The catalog comes first.

Where to start

Make a free account — it does not ask for a credit card — download the helper from the Sources page, and run sally go against the drive you have been avoiding. You can have a card catalog of your whole mess by dinner, while every one of your files stays exactly where it is. That is the whole offer: I would rather know your files than keep them.

Start free at sallydoesdata.com/signup