High Output File Management


The best way I've found to organize my files is to split them between "created by me" versus "created by strangers". If I write an article, it goes in one place. If somebody else wrote it, it goes someplace else.

This distinction is important, but it is also fuzzy. Plenty of photos of my family were not taken by me, but I store them in the "me" pile. I have an install directory designed for recreating my setup, full of drivers and profiles and so forth, and even though I didn't create most of that, it represents a configuration I did create. What about bookmarks? Bookmarks can show up in either pile, no big deal.

"Did I make this or not?" really seems to be the highest-order bit for organization. Old PDFs of my LaTeX homework solutions are completely different from a magazine article I downloaded in PDF. They don't belong together in the same folder.

In fact, it seems that the things I worked on are my only important files. Everything else is just cache. Out of ten top-level directories, only one is for non-personal files, found. That's where I store the infographics and other miscellanea from the internet. If I have to think twice about deleting it, it doesn't belong in found.

For a long time now, the internet has had a sufficient number of people volunteering as archivists that the rest of us can take a breather when it comes to things like "rare B-sides" and "fan movies". I only archive things I need that no one else mirrors.


You really only need the last three years or so of email online. I have offline archives going back about twenty years, and I have never once gone back to read old email. The stuff that's worth saving gets plucked anyway. I still can't justify deleting the emails—there's too much personal correspondence mixed in—so I compromise by moving them offline.

When Gmail first appeared, Google encouraged users to stop pruning email. But they were wrong about that. It isn't enough to simply have the capacity. The downside to not pruning is that eventually your email becomes impossible to search. Real deletes help that. I now try to trash anything automated, impersonal, recurring, or replicated. Letters that other people have written me are high value. Shipment IDs are already saved on Amazon.com.


We save the things we make forever, but most of the stuff we post online has a limited shelf-life, at least when it comes to outside appreciation. In practical terms, the only time people ever read through old tweets is when they're looking for something damaging. Which means they don't need access to them.

Initially, Snapchat sounded absurd. How could you stop anybody from taking screenshots? But the system works, not because it stops people from saving any one snap, but because it makes it prohibitive for anyone to keep a running archive of all snaps. The archive is where the liability is.