Clean up your desktop with this one weird trick

Desktop clutter is so 20th-century. That was when I first started dealing with it—1991, to be specific—and for decades, it was something I simply learned to work around. What I needed was a digital butler to quietly clean up after me.

Up until recently, my desktop was a swamp. It was only too easy to use as a dumping grounds for temporary stuff, notes, images, and other flotsam I hadn’t filed.

This was not only embarrassing during meetings, when other people might be looking at my screen, but—similar to living in a cluttered apartment—it was also demoralizing. Starting my day staring at a screenful of chaos did not put me in the best frame of mind.

Typically, I’d work like this: let files accumulate, and then, when I couldn’t stand it any more, create a folder named “clean” and throw everything into it. Bravo! Nice clean desktop. Then the cycle would repeat itself. Some of my backup hard drives have nested “clean” folders about eight deep. No, this wasn’t good.

An app called Hazel, by Noodlesoft, changed everything. Hazel basically watches any location on your computer you want it to—desktop, Downloads folder, etc—and can run automated operations (rules) on the files in those locations. Short and sweet, here’s a few of the things Hazel rules can do:

  • Move all image files over two days old to a specific folder
  • Unpack zip files, and continue to run operations on the contents
  • Trash disk images an hour after you’ve downloaded them
  • Grab all files starting with a specific word

Best of all, the interface is very similar to the native Mac UI for smart folder criteria. I’ve encountered organizational apps that, while powerful and versatile, were saddled with a UI that only made sense to the programmer who designed it. Not so Hazel; if you’re familiar with the Mac search-critera UI, you already know how to set up Hazel rules.

Hazel runs in the background, conveniently accessible via the menu bar. Typically, you won’t need to manually cue any of the rules, since Hazel does that every couple of seconds, but if you’ve made some changes and don’t want to wait for Hazel to run, you can do it via the menu icon:

Hazel also allows you to synchronize rules across multiple machines via Dropbox. This is fantastic.

Cleaning up the mess

Here’s what I did.

First, I took a closer look at the stuff that accumulates on my desktop (and Downloads folder, which is even worse). Much of it fell into the following categories:

  • images
  • saved webpages
  • Word docs
  • disk images
  • zip files

But I also noticed that the majority of these items were text files, named things like “idea 1.txt”, “todo sdfasfsad.txt”, and “notes3.txt”. This is because I spend about 60% of my time in a code editor (Panic Inc’s excellent Coda, for anyone who’s curious). If I’m in a meeting, the easiest thing to do is start a new file and take notes there. If I have an idea for an article, I’ll typically start writing that in Coda as well. Over time, I developed a workflow where I’d use Coda as an informal notepad, saving each file to the desktop and prefixing the filename with a brief descriptive tag.

The fact that I’m meticulous enough to at least classify my clutter made setting up Hazel rules pretty easy. I now have a series of rules that can identify and file the following text files:

  • project ideas
  • meeting notes
  • journal entries
  • useful snippets of code
  • temporary scratch-pad stuff (source code I’m cleaning up, clipboard text, etc)

So I can continue to use Coda as a notepad, and—as long as I’m good about prefixing the filename with the right tag—Hazel files everything for me. I work the way I work, and things get magically put in their places; no more post-apocalyptic desktop. Nice!

Yeah, but…text files?

There’s a case to be made that maybe this isn’t a fix so much as a band-aid on a flawed process. I’ll admit something seems a little hacky about using one’s IDE as an all-purpose notepad, and relying on an organizational scheme that’s basically the digital version of putting sticky notes in file folders.

As it is, I already have a place to put a lot of these files: DEVONthink, by Devon Technologies. I keep DEVONthink databases of pretty much everything I work on—Vassar projects, personal projects, ideas, etc. If I were as assiduous as I’d like to be, I wouldn’t need a computer program to file my random scraps of paper, because I wouldn’t have random scraps of paper—I’d be putting these items in DEVONthink where they belong.

At issue here is the amount of effort required for each interaction.


  • Switch to Coda, assuming I’m not already in it (I usually am)
  • Start a new file in whatever window is open
  • Type, save.


  • Switch to DEVONthink
  • Open the specific database (a project, “Future Ideas”, “Blog Posts”, etc)
  • Navigate the internal file structure of the database until I’m in the “meeting notes” folder, the “articles in progress” folder, etc
  • Create a new file
  • Type, save

While the DEVONthink UI is pretty clean and straightforward, filing things directly in DEVONthink still means more fiddling than I’d like. The other issue is that a lot of times, the items I have Hazel auto-file and the items I’d store in DEVONthink are different types of files. Temporary text files, off-the-cuff todo lists, and disparate ideas are general-purpose enough that I don’t have a DEVONthink database for them. For items that should ultimately wind up in DEVONthink—like meeting notes, which I keep in each project’s DEVONthink database—filing them with Hazel makes them easier to find later on: I can take notes in the meeting, and then retrieve them afterwards.

So for the most part, I’d say my methods are complementary: I have DEVONthink for the important stuff that needs to be archived in specific locations, and Hazel for the simpler stuff that just needs to go where I could find it if need be.

Desire path

The other way to look at it is that my Hazel-based system of classification is essentially a desire path—an organically-evolved way of working that, for my purposes, is the most efficient way of doing what I need to do. (Desire paths are actually pretty instrumental to design—although they may not always be paths.)

I’ve spent most of my professional life experimenting with organization tools. In college, I had a sticky-note-based scheme. The first task-management app I used was Microsoft Entourage’s rudimentary project-management tool. I’ve used BaseCamp, Trello, TeuxDeux, various hacked-together WordPress solutions, and—currently—Asana. (I even used a local installation of MediaWiki for a while—that was an interesting experience.)

The number one thing I’ve learned: Find tools that work the way you do. The core to staying organized is rooted in habits, not software. Apps only solve your problems when they make your existing way of working more effective; organizational schemes that require intrusive/counter-intuitive ways of working won’t last. If I wasn’t in the habit of prefixing my text files with “note” or “project-idea”, I’d still have a desktop full of text files. Hazel didn’t require me to fundamentally change the way I take notes; it just significantly reduced the overhead of keeping track of them.

I don’t know if it’ll work for you, but if digital clutter is a problem for you, Hazel might be worth a try.

(Disclaimer: I have no affiliation with Hazel, or the people who made it—I just think it’s a great app that solved my problem.)