Interestingly I have been following a yyyymmdd-major cat-minor cat-subject.ext format for over 10 years. My folder structure is date based on 2020>QTR1…QTR4 (to reduce the number of documents in one folder). Even without Foxtrot Pro search (which is excellent) I can pretty much find anything I need in seconds. I use hazel to rename the document and file it in the archive. I work from one 30-Day folder (so named) in which all my current working documents reside. As soon as any document hits 5 weeks with no modification, Hazel files it. If I need to repurpose content, I always copy the document and it gets a new date. The long file names can be a pain on mobile devices, but in general I have a large screen. This also allows for moving archives off my system for long term storage when they are 10 years old. There are a couple of exceptions to this: videos/audio and photos are in a separate archive (Lightroom) and teaching material has a topical folder structure.
I tried this bracketing system with a subset of files. I was using bulk renaming tools which helped, but the tool I was using had bugs and I ended up with a lot of skipped items. This sent me on a quest to reorganize everything, also motivated by the recent Notion outage and finally deciding (once and for all?) that I am done with any software that doesn’t function if the internet is disconnected.
I ended up with a small set of folders, based on areas -
- Do - Hobbies - birding, genealogy, learning, writing, etc.
- Maintain - Fitness/Health and vehicles
- Refer - reference material - financial records, receipts, manuals, recipes, etc.
- Store (archives)
- ZK (zettelkasten, PKM-type stuff)
Not sure this structure or naming system will stick, but it’s a start and I at least feel more organized, so far…
AAaaannndd…it didn’t.
Shout out to @pantulis for the link to Johnny Decimal - an incredibly easy system to switch to. I’m still surprised it was that quick! And it feels right for my brain.
This is one of my pet… obsessions? File organization, information management systems, I’ve been thinking about this for decades, and went to grad school for HCI. I’ve had three main systems over the years:
-
File/Folder hierarchy based on topic - This is the most common method used by most people. The only trouble here is that files tend to get named randomly, will be misfiled, or there will be files that fit into two or more categories but they can only exist in one folder without setting up aliases… which no one does.
-
The Bunsen Method - So named after a system I read about on the Dr. Bunsen blog, now defunct, where you use a shallow hierarchy, like everything in Documents, and use a rigid filename system to categorize files. For example,
2021-03-05_a_b_name-of-file.pdf, where a is a category, and b is a sub-category. This works well until you have hundreds or thousands of files, then I’ve found that Finder tends to struggle a bit with the folder. Also, you have to be very rigid about the naming system, or it all falls apart into chaos. -
Finder Tags and Shallow Folders - This is what I’m using right now. Hazel tags everything, renames everything, and then moves the files into a sub-folder of Documents named for the type of file that it is. So, a pdf goes to the
~/Documents/PDFfolder. A Pages document goes to~/Documents/Pages, etc… So far so good. Files are named with the date and what they contain, like2021-03-05_letter from aunt cathy.pdf, and tagged with a flexible hierarchy like “family”, “personal”, “extended”, “letters”, or whatever.
The purpose of any of these systems is to think about the system as little as possible, but trust that whatever you put into it, you’ll be able to quickly and efficiently get out of it. Everyone’s brain works a little differently, so you have to choose the system that works best for you.
My biggest concern is that Apple doesn’t do enough in macOS to help manage the deluge of information yet. I’d like to see:
- Leverage the AI of Siri to suggest similar documents on demand in the “Today View” sidebar thing you swipe over from the right hand side of the screen, similar to DEVONthink.
- Universal tagging. Finder tags in Reminders, Mail, Calendar, Maps and a third-party API so every other application can take advantage of a single data organization system.
Unfortunately no one at Apple seems to be working on anything like this. Ah well, maybe someday.
I really like the Johnny Decimal system and some of the variations you’ve suggested. But after spending way too much time trying all kinds of systems, I realized something. I’m a simple guy, I need a simple system. So for the last 6+ months I’ve been using this in iCloud Drive: (I’m an iPad only guy)
- Desktop
- A folder for each active project (‘2020 Taxes’, etc)
- Documents
- A folder for each category (i.e. ‘Library’ for my ebooks, ‘Drawings’ for drawings I’ve made, ‘Medical Bills’, ‘Finances’)
- Downloads
- anything I download goes here first, I clean it out on Sundays
Its a shallow file structure, and each folder has just files, typically “YYYY-MM-DD descriptive name” format. When I’m no longer working in a folder on my desktop, I move it to my Documents. Occasionally I move stuff I don’t use to an external hard drive.
I use a similar system at work, just mostly residing within OneNote/shared directories. (Windows 10 and approved software only)
I can imagine this not scaling well for people with more involved needs than me. But as one of my coworkers told me “The job of an engineer isn’t to add as many features as we can. It’s to get rid of everything we don’t need”.
I used the Bunsen method for a few years, but when I joined project teams that had large libraries it became too time consuming. I’m comfortable now with project folders and subfolders for tasks and for categories – in DEVONthink. I use DEVONthink’s replication feature. lot. E.g., put a presentation into /my_project/2020 Presentations and replicate it to /my_project/Sponsor Meetings/2020/12. Other than date prefixes, the rest of the file is a short string indicating content.
I find Finder tags and DEVONthink tags are far too time consuming to manage and do not want to spend the time disciplining a tag structure. Finder tags in particular – it seems to be impossible to keep stray, irrelevant tags from creeping into Finder tags.