aaron's picture

Hard to clean spots, No. 8: My Work folder


My Work folder is for the documents that I (or any collaborators) produce for specific projects. The important organizational idea here is just identifying the area of my life to which these documents chiefly pertain.

This used to be a lot worse than it is now. On its way now, but there are still some big question marks.

Mindful Media Management. Work Folder

As mentioned in a previous post, you find my Work folder in ~User/Documents. Inside, my Work folder is broadly segregated into context areas: AS, clients, education, GRS, employment, projects, MRB.

Ideally, each folder inside Work is flat (e.g. has no subfolders) as close to the root folder as possible. The folders that come closest to this ideal are my personal folders (named with the initials of my family members). AS GRS and MRB are basically flat.

Some holdovers from the heyday of folder-filing-cabinet-inspired filing do exist, though. These still need to be flattened out. One day, maybe.

I have previously flattened deeply nested folders by embedding each element (e.g. folder name) of the directory path into the filename (using the human-readable separator string ". ") of each file in a given folder.

Mindful Media Management. AS Personal Folder

So instead of having to click through many levels of folder hierarchy, you navigate to the 'root' folder and scroll thru a list of human-readable 'tags' or 'contexts' to find a file manually, or search using GoogleDesktop or Spotlight using the 'tags,' 'keywords,' or 'contexts' embedded in each file name.

Here's how I converted from deeply-nested to flat folder hierarchies: I selected all the contents of a folder and auto-renamed all selected files with the name of the folder they were in. Then I moved all the files in that folder up one level, deleting the empty folder and repeating the workflow until all deeply nested files were as close to the root as possible.

Particularly in my personal folder, [AS], the order of file name elements has a specific meaning and importance. Note: This is a work in progress that I'm still struggling to describe with a (simple) rule. But here's how it stands now:

Everything in my AS folder is prefixed with AS. Then, different contexts of my personal life are listed and each is followed by with their most immediately important sub-contexts; at the beginning of the filename.

So when naming a file, I leave the original filename at the end, to which I append my unique descriptor for the file (before the original filename). And the AS + Context + Subcontext goes at the beginning.

Context [Subcontext]

Business Owner [Consultant, Reseller, Vendor],

Commuter [NYC MTA, Translink],

Consumer [Air Miles, Bills, Deal Hunter, Event Registrations, Invoices, Legal, Menus, Research, Returns, Scams, Serials, Travel, Wishlist],

Domicile [Real Estate, Renter, Moving]

Driver [Directions, Licensing]

Education [Grad School, PNNL, SFU, U of L, UBC],

Employment [Application, Co-op, Job Description, References, Résumé, SFU]

Finances [Banking, Student Loans, Medical Bills, Taxes]

Healthcare [Benefits, Coverage Networks, Forms]

Identity [Citizen, Security Clearance, Metis Citizenship]

Personal [Legal, Values, Wedding]

Research [Troubleshooting]

Writings [Journal, Notes, Letters, Poems]

Near the end of this list, this schema gets kind of muddy. For example: the whole AS folder is my personal folder, so having a 'Personal' context seems kind of moot...Some of this remains to be worked out. ('Personal' might, for example, be dumped into my Identity context...or not.)

Another problem: I originally approached these contexts and subcontexts with the idea that everything could be reduced to a role in my personal life. But certain word forms for 'role' are quite wordy (e.g. Banking is more succinct than 'Bank Account Holder'). And some subcontexts are actually about the type of document in a given context. (Consumer subcontexts are a good example of this.)

UPDATE Saturday, September 12, 2009 3:29:43 PM

Okay, so when I originally wrote this list, my Work folder made it for a couple reasons.

First, to revisit some explanation: I have my Work folder split into different Work contexts. Each work context has its own folder, which usually contains more subfolders.

For example: in the folder named 'education' for my work done in the 'education context,' I previously had two main subfolders: undergrad and grad.

Since I have been through several graduate programs, I have a folder for each one.

Mindful Media Management. Education subfolders

I used to have to go to 'Documents > work > education > grad > SIAT > Course Folder' to find or file my current coursework. Thats 5 clicks just to get to a directory full of folders named for each class. So 6 or 7 clicks to find or file coursework, given any subfolders. (I solved the 'deeply buried folder'-problem initially by having a shortcut to each course folder on my desktop, and then switched to a stack full of aliases pointing me to my 'most-used-places.' This alias stack is pruned frequently to nuke aliases to folders I no longer need frequent access to.)

Mindful Media Management. Alias Stack

But, it has been many years now since my undergrad work was completed (or even accessed) and even my earlier grad work is years old now. So most of those files have long since been archived.

So even after I archived (moved) my undergrad and earlier grad work to an external drive, the directory structure that made sense years ago remained... Yet today, all this directory structure did for me was make it 1 click 'harder' to get to my school work.

So I recently removed the directory that corresponds to a now-useless distinction (the old Grad folder is now gone).

This is a simple example of bringing all folders as close to the 'root' as possible.

This example also highlights the big problem: my clients, education, employment, and projects folders each have a number of subfolders, which likely have their own subfolders. Whereas my goal is (i guess) a flat folder structure: most of my work context folders are not flat.

The problem is finding a method of auditing the utility of conceptual distinctions used to name folders that rewards shallow hierarchies.

This is principle is negotiated slightly differently for each work context, given the context structure and constraints.

Mindful Media Management. Work Subfolders. Education folder

My work > education folder.

Mindful Media Management. Work Subfolders. Education Subfolders. SIAT folder

For example, the work > education > SIAT subfolder segregates courses I've taken from those that I have simply researched; my course data, and administrative documents. (This could now be far more easily done with Tags; though this view is far cleaner compared to the VAST Lab view below, so there is a tradeoff in using folders to segregate files, rather than filename elements.)

Mindful Media Management. Work Subfolders. Employment folder

My work > employment folder.

Mindful Media Management. Work Subfolders. Employment Subfolders. VAST Lab folder

My work > employment > VAST Lab folder is nearly perfectly flat. When sorted by 'Name', the subcontexts emerge as you visually scan & scroll down thru the folder contents...

Mindful Media Management. Work Subfolders. Projects folder

My work > projects folder.

Mindful Media Management. Work Subfolders. Projects Subfolders. Text Vis

My work > projects > text vis folder has subfolders named for stages in data processing.

Different projects have completely different subfolder structures, based on what makes sense for the given project. Not all of these subfolders have coherently named contents, though, so that is something to work on.   bookmark tuals 0.1 on

Bg Bottom