Johnny.Decimal
johnnydecimal.com
johnnydecimal.com
Edit: yep. As suspected.
Let me just say this, if I may:
1. You don’t have to like every idea, but please consider that some people do.
2. I’ve been using this successfully for a decade. I’m using it right now to organise a project. My boss loves me for it, because guess what? We can find things!
3. I’m not alone. I’ve done essentially nothing to promote this yet still I get multiple emails a week from strangers telling me that they love it and asking for guidance.
Okay, carry on. I’ll answer questions here where doing so feels productive.
if on the other hand i see something that's problematic or could be improved, i should say that, instead of just downvoting.
i think the problem comes from the tone people use to voice their disagreements
On GP's:
> My guess is the cynics are the ones who usually comment while people who are more positive will probably be more likely to just upvote?
It irks me to see this comment pattern regularly being called out as "HN specific", because it isn't. It's universal in on-line discussions - particularly those on platforms with upvote/reaction buttons. There is nothing interesting in "I like it" comments, so you always get disproportionately more of critical comments ("I don't like it because", "I had a problem with it"), tangents, and low-quality humor. HN is perhaps different than most places in its strong dislike of low-effort jokes - which leaves us with just criticism and tangents. My point thus is: it's not HN, it's the Internet. It's not a mutual admiration society like you get when you go out with friends for a beer. I would think people would've gotten used to it in 2020.
> It irks me to see this comment pattern regularly being called out as "HN specific"
I've heard this general rule vocalized on social media as "Fans like, haters comment".
If you get loads of upvotes for an idea and a lot of flack, those pushing back is not your audience and can safely be ignored. You are not doing anyone any favors by trying to appease everyone.
I’ll say that Tailwind CSS is a godsend for someone like me. Being able to semantically describe what a thing should look like, inline, is miraculous.
Edit: /s/tailscale/tailwind. Tailscale is also amazing.
I imagine it's a minority view here (well, pretty much everywhere), but I relish getting constructive criticism of my works, whether it's on a site like this or a review by my peers/customers/employers.
That's because I have my own perspective, point of view and view of the world. Others have different ones.
When other folks share their opinions/perspective, it gives me an opportunity to examine my own preconceptions -- which usually winds up giving me a better understanding of the topic at hand.
Granted, in a setting like HN, there will be those who will be uncharitable or dismissive, and those voices can be ignored.
However, those who criticize specific stuff and provide the assumptions/reasoning/thought processes behind those criticisms can provide valuable insight into how you and/or your product/service are perceived outside of your small circle of shared views and assumptions.
I hope the bulk of the comments here are of the latter kind rather than the former, and that you get new perspectives that can drive positive growth for your activities.
...sometimes. ;-)
I use the "feedback bucket" mental trick: all feedback is shit (whether supportive or critical). All the feedback goes in the bucket, where it nourishes the growth of the beautiful roses of customer insight.
The point is to remove the immediacy of the feedback, detach it from the thing (and my feelings about the thing). But still allow it to inform my thinking about the thing. Instead of taking a criticism to heart, I can mentally "put it in the bucket". If it's useful, thoughtful, considered, constructive criticism then it will produce some insight later. If it's just spite, then it'll produce nothing.
> If you choose to give me your email address, I will treat it with the utmost respect. I’ll never spam you, or give it to anyone else.
> In 2019 I moved from Mailchimp, whose business model started to make me uncomfortable, to Buttondown. It’s run by one guy who I trust to do the right thing.
If that privacy policy doesn’t make it screamingly obvious that I’m not tracking you in any way then holy moly I don’t know what to do.
I think if[1] I ever had a website, and if[2] I had advertising I would choose one advertiser per month, display a static image at the top of each page, and clicking on it would take you to a static page on an advertiser's website. and it would have a privacy policy like your good example
[1] big if
[2] bigger if
Maaaaaybeee you might want to strip last octet in IP number from logs, but that's pretty much it.
Complex GDPR policies are only necessary when you want to store and process Personal Data.
It’s hosted on Netlify, so I don’t even have logs. :-)
GDPR compliance is overblown, when the simplest way to be compliant is to "just don't store nor process personal data"
The idea intrigue me! It's not often you see kind of thing that seems mundane(to general interwebz population) like a system to organizing files and folder. There is some method floating around, but they either too basic or too complex for my need.
That said, back then I give Johnny Decimal a try a bit, but works took toll on me and I eventually give in to my old way of filing things. Maybe it's time to revisited. :)
Website is clear cut and to the point, its gorgeous btw. :-)
My first reaction was creeping horror, because this feels like a solution to a problem people shouldn't have. But now that I read you talking about the situations you've made this for, I get your point: it's a problem people do have, and one that's not going away any time soon.
Just to save future bathub wear and tear, I'd suggest adding a bit to the top of the website to say who this is for. Just a couple sentences telling the story of your archetypal user would do it. You might also say who it probably isn't for, just so we can get to the dogs wearing clothes images sooner.
“It isn’t for everyone” is absolutely right, and if I can make that separation higher up then it’s better for everyone.
There being new filing / organizing systems and apps (zettelkasten is the new hip thing nowadays it seems) popping up on HN every once in a while isn't helping either.
Tl;dr, I think it's the idealization of organizing, the feeling of productivity, etc that is fueling a lot of people.
Meanwhile I've gone through three or four different note keeping / taking / managing apps and still haven't settled on anything, lol. I also lost the need for it, now that I've gone from the exploratory phase to actually building software.
Not suggesting it though :P
... or you drown under the ocean of unrelated finds, which is my usual experience with mass-applied windows indexing (in the form of massive corporate sharepoint)
Do I introduce another decimal point? Like so: "1.123.0"
My top-priority job as soon as I get time over Xmas is to get (an updated, better) version of this up on the main site. It’s the #1 question I get asked.
Took me a few years to figure this one out but I’ve been running it in anger now for a year and I can say that it works.
Second, have you experienced any of the problems outlined in MetaCrap https://people.well.com/user/doctorow/metacrap.htm
How have I never seen that link?! Yeah the issues described therein sound familiar...
I have one question: is there JD best practice on how to handle file sets that are generated frequently and in large-ish numbers? I’m thinking specifically of camera image files. My current approach is to have a directory per year, and then subdirectories (with a YYYMMDD-subject name) for each distinct shoot (typically between 30 and 50) in that year, together containing between 5000 and 20000 new images (depending on how many shoots I do). I maintain a parallel folder structure for the processed images. My current thinking is to keep this photo-library as-is, but put all of my ‘normal’ files into a JD structure.
By the way, and picking up memetherapy’s comment, I’ve also experienced knowledge management problems in large corporations for which JD would be a great answer. The last place I worked moved several years ago from a ‘classic’ system based on servers and file systems to a system based upon numerous per-project sharepoint sites. It didn’t help that the template for the sites was developed from a project lifecycle perspective and had about 70 folders covering every stage from initiating a bid to project shutdown. There was pretty-much a single ‘dev’ folder for the actual work, that techies like myself inevitably populated in inconsistent ways. Something that was never resolved was that all of these silos made searching for old reports, etc really really hard and fundamentally broke knowledge management in the company. Especially when most of the silos had per-project access permissions.
On the point made elsewhere about tagging, in my experience it is extremely hard to get people across a large company to add metadata to content, let alone do it consistently.
Yes. This sort of data set already sorts itself pretty well: by date. There’s no need to mess with that.
See also: your music library, if you don’t have iTunes do it for you.
> It didn’t help that the template for the sites was developed from a project lifecycle perspective and had about 70 folders covering every stage from initiating a bid to project shutdown.
Aah, when the PMO tries to be helpful. I see this all the time - a ‘baseline’ set of folders that you’re meant to understand know what to do with but literally nobody does so they just get ignored.
Worse, I frequently see them with numbers at the beginning. Worse even, those numbers just make no sense at all. I’ve recently experienced a repeating pattern which involves folders 001-015 being something that I can’t remember, and then `016 TO BE DELETED` (because this is in a document management system where you can’t actually delete a file).
Given how the system is setup, I find this amusing in a meta-way.
Also on my to-do list: get a few example structures up there. But drop me a mail and I’ll send you my home structure for inspiration. hello <at> the domain.
I'm gonna organize my 2TB thumb drive this way tomorrow (I'll at least start) and see how it goes...
Elsewhere someone suggested going 36-bit. Or hex, or whatever. So after 99 you’d go A0 and so on.
I don’t love that — the neatness gene is making me itch — but it’d work.
I’d say perhaps you’re being too granular if you have 100+ of a thing in a category. I’d love to know what it is, if you want to mail me. All this info adds to the system.
If there is an obvious delineation, I will accept it and use it. If there is a line separating categories and that line is blurry at all, I find it impossible to both confidently choose where it goes and know where I filed it later on. I can know where to put something or I can know where to pull it from. Never both, somehow. (I am as confused by this as you are, but it is true for me.)
So everything winds up in piles, both physically in real life and in my filesystems. The only way I can remember where something is, is to remember what the pile looked like when I placed that thing on the pile. Then I know exactly where it is.
In network file folders I can find something by remembering my own history in that folder and remembering about how many items were there when I created the thing I'm looking for. Then I sort by creation time and I count that many down, and it's usually one or two away from that number, in one direction or the other.
My windows desktop is organized autobiographically as well. I can never remember what I named a folder, but I know exactly where it is on the desktop.
I'm basically a nightmare for anyone that needs a tidy environment. If you clean up my stuff, everything you organized is lost to me forever.
After reading this post, I have started to organize my documents and projects according to the description. I feel like this is going to be helpful not only so that I can find what I'm looking for, but so that I don't have 4 copies of every giant source code repo taking up my disk space.
I do have one question: in the examples, the category shows "40-49 BlahBlah", but the subcategory goes "41 Taxes, 42 Expenses, etc." Is there a reason why X0 is skipped? Is this so you have room for something you might have forgotten?
> We reserve these numbers for ‘management & meta’ information about the category. Although I hardly ever use them, it’s rare that I have more than nine categories in an area. I don’t miss the wasted number, but I always appreciate it when I need it.
Also recently I’ve been starting my IDs at `.11`. Same idea — sometimes you really want stuff to always appear ‘at the front’ and this lets you achieve that.
In the project system I’m setting up now, for example, we’re using the PMP methodology [0]. This very neatly divides up in to:
21 Scope
22 Schedule
23 Cost
...
29 Stakeholder
Then I won’t fill up the zeros, I’ll skip straight to .11 in each category for the things that are ‘personal’ to this particular project. I’m expecting that they will not be standard across projects.
[0]: https://www.pmi.org/certifications/project-management-pmp
Sharepoint should have been able to fix all that, but somehow it just made it the same, but slower. Even after 20 years, I never know where the f__k I am in sharepoint, or when someone sends me a link to a sharepoint doc, I can open the doc, but I don't know where the f__k the document _is_. I dont know quite why sharepoint is so confusing, I think its because it tries hard to pretend to be something other than the document store that it should so obviously be. Just when I get the hang of it, they redesign it all, and add some new front end like Delve.
If I join a project, something that rings alarm bells is if the Project Manager has created sixty empty folders in Sharepoint arranged into a three level hierarchy with different folders for every stage and stream of the project. Most of the time, most of the folders remain empty for the duration.
So yep I really like this Johnny.Decimal idea, simple and workable.
Seems to be no different than this:
>something that rings alarm bells is if the Project Manager has created sixty empty folders in Sharepoint arranged into a three level hierarchy
Except that it's a hundred empty folders in a two level hierarchy replicated by convention across different systems, including Sharepoint if your organization uses it.
This. One hundred times this.
Those here arguing that things can, in fact, be found, or saying that my claim that "nobody can find anything any more" is absurd, clearly haven't worked in the ~5 large enterprises that I've worked in over the last decade.
Internal organisation in enterprises is shockingly bad. Remember, these people aren't techs. They have no idea what `find / -iname file | grep whatever` means. They crack open SharePoint and rummage around. When they want to save a thing, they just put it wherever. It's chaos.
I'm just implementing this for a few new projects at my current place and my boss has messaged me a number of times expressing his joy at just knowing where stuff is.
Another thing for Johnny: You may not need to get the IT team to make you a Sharepoint list for tracking your IDs. You may have your own little Sharepoint site if they use MS365. OneDrive for MS365 business runs on Sharepoint. So just make your own there.
Admittedly, I don't, because I'm actually using OneNote for that, as well as storing the notes. So every Area is a section group, Category is a section, and every ID is a page. And if I need to make more notes in an ID, I can make a subpage. But normally it's group-section-page. This leads into: the default page is 00.00, which is the table of contents. If you want to go really crazy, you could just embed any files into the pages as well, but I currently think that's a kinda nasty way to do it.
A third thing: if somehow you end up with more than 9 categories (see prior note on categories corresponding to people), nothing stops you from using Base36 for the category numbers. As above, if I became a middle manager, and had to track more than just a few guys, I'd still continue with lexicographic sorting category "numbers". It'd just be non-numeric. This goes into breaking the rules, but rules are there to be broke when necessary and sensible.
This system started when I ran a dance production. We were touring to multiple venues in multiple states, but they all have ticketing, seating plans, merch sales, etc.
I can still remember that system and I literally haven’t looked at it in ten years.
- 40-49 Venues
- 41 Melbourne, VIC
- 41.01 Tickets
- 41.02 Seating plan
- ...etc.
- 42 Sydney, NSW
- 42.01 Tickets
- etc...
$ CDPATH=$(find /path/to/cjd/* -maxdepth 0 -type d -print0 | xargs -0 printf "%s:")
$ cd 34[tab]
$ cd 34\ Events\34.[tab]
34.18 Product Launch
34.19 Company All-Hands
34.20 Holiday Party
$ cd 34\ Events\34.18[tab]
$ cd 34\ Events\34.18\ Product\ Launch
Happy to link it back to this comment or any other profile you send my way.
You can do a simpler version and just set
CDPATH=/path/to/cjd
Or you could change the path in the find command to
/path/to/cjd/*/*
- It only solves the problem of communicating categories to another person/party with its brevity. 12.03, easy to tell others. But all this at the expense of the following things.
- There is no way to immediately know what 12.03 means. One has to carry a look up table. You can replace it with 2 char code. For example, Finance.Taxes to FN.TX. Why use numbers? Is the ordering important? Why not use char?
- What if categories do overlap? GMail solved this problem in 2005 with labels instead of folders. You can apply multiple tags. Tagging should be enum-like, while tagging it should show what tags already exist so that you dont create two seperate but similar tags, for e.g. Finance and Finances.
- The author dismisses search without giving any reason. Search is amazing for digital documents that is not possible in physical analog documents. The focus should be on tags + keywords or description of the document. Then use something like Algolia search to find stuff - it is extremely forgiving and powerful. Ofcourse, I am just speaking in general about data organization, not specifically about OS-level folders. For that, I think we're stuck with what the OS provides as a search engine. On MacOS, I think Alfred does a better job of searching, haven't used it personally though.
- You can still assign a unique ID to the document, after all thats all the 12.03 scheme does. You can communicate precisely to Jane, "Hey Jane, the document is B75AE2". Jane types that in the search engine and there is no need to weed through folders.
- Additional metadata such as year, author, owner, etc. would help with search.
Instead of making search powerful and contextualizing it, the author expends his arguments on frivolous pursuit of Johnny.Decimal.
You went on to mention labels and tags as a solution there -- but the point here is to implement the structure in a common filing system, isn't it? Your point works fine for email, but for documents on my drive tags don't seem like a great approach because at least on my machine even though tagging is an option, it's not robust and reliable.
It didn't feel totally fair to say that the author needed to address search more; the ability to search doesn't go away when you implement a system like this, it's still a powerful tool available to you if you know what you're searching for. If you don't know, then a system like this is a good way to explore the topics and assets in your file tree in a way that allows you to discover collections of related files.
Lastly, true that you can assign a unique ID -- but again, isn't that introducing and solving a different problem? Uniquely addressing every file might be a really useful thing, but as you mentioned, there's no context clue to help you decode B75AE2 -- you would need a lookup table. Even then, presumably you would want to know which categories and families that file should belong to.
Overall, everything you listed here is thought provoking and I think points out some of the limitations of the way our digital file systems emulate physical ones, but this collection of objections isn't persuasive that the Johnny decimal system isn't a good idea.
I'm not sure it even works there. I have emails both categorized and tagged and still can't find anything between the thousands of work emails I get a week from coworkers, mailing lists, and automated systems. Any search brings up dozens to thousands of results.
> but for documents on my drive tags don't seem like a great approach because at least on my machine even though tagging is an option, it's not robust and reliable.
Not to mention that most users aren't interested in tagging every file they have. I never tag files. My emails can get auto-tagged by mail rules, but I don't have time to add tags to all of my files.
What do you mean? How do you expect the author to "make search powerful"? This system is clearly in the context of a desktop/NAS filesystem, so they're stuck with whatever search systems exist. I honestly don't plan to use this system, and still feel the need to respond to this critique of it.
> Why use numbers?
Is answered here[0], to intentionally preserve ordering. You don't need to know what a plain ID means, because the only reason to use it is in context.
> What if categories do overlap?
There's nothing to stop you from tagging things in addition to using this system, if you want.
> dismisses search without giving any reason
Again, the author is not building their own document storage system.
> Additional metadata
This system is trying to solve the problem of having to supply a bunch of information to find a specific document.
Overall, there are lots of anecdotes here about how this system is useful in some situations. There's really no reason to sound so incredulous about it, even if it's not for you.
- Discoverability: How many times have you opened glossary of a book just to explore and discover what else is in this book? Or open up a telephone directory with no specific person or business in mind, but a general category - "Shoe repair"? I think when things are organized by category, it allows for discoverability which would be difficult in a flat-searchable structure. Search works great when you know what you're looking for.
This can be half-solved by listing all items in a tag. Just that they'd appear in multiple tag archives as duplicates.
I've tried this (in the context of multiple projects [0]). Believe me, I've tried everything!
The reason the JD site is the way it is is that I've been doing this myself for a decade and this is what I consider to be The Best Way.
It's why we have street numbers to find buildings. It's why libraries use catalogs to locate books. It's why financial systems use a chart of accounts with numbers and prefixes to organize transactions.
Location is important, and abstracting that away into a giant file share with search only makes things worse.
These days I hardly ever bring up file naming in code reviews—fzf just makes it so easy to find everything.
You can hit CTRL+t from the terminal after installing FZF and it will let you fuzzy search for any file you have. So you can do something like run `vim [CTRL+t]` to quickly find and open any file you want, or `cd [CTRL+t]` to switch into some deeply nested directory.
Even inside of WSL 2 with a 6 year old workstation on a first generation SSD it takes ~3 seconds to index 190,000 files and once that index is built, narrowing down results while typing is close to instant when using ripgrep as FZF's search back-end.
The amazing thing about all of this is it's dynamic. There's no having to create aliases or pre-defined directory structures. I only discovered CTRL+t from FZF a few weeks ago but once I did it was almost as good as discovering CTRL+r for the first time.
I search "driveway" and get all the photos my wife and I ever took that shows our driveway. Ah yes the cracks are getting worse.
I search email for "driveway repair" and find that old recommendation from a friend despite the word driveway having never been mentioned.
I don't need a sorting system. I need a big pile of all my data ever with a fantastic search engine.
I'd prefer if it wasn't google.
The problem is when search extends across domains. I can search files with Spotlight, but not my google emails. On my phone, it's worse because there are so many different comms systems. I can search my Whatsapp messages; my SMS messages; my email; Facebook messenger; Skype.... Each individual search is fine, but no one tool does them all. Hence, the idea of using manual labels to classify stuff - including emails - can still be useful. Crazy that we're forced back to this.
Edit: Seva and Cloudtenna
I wish I had a local one that just packed it all into Solr
Yep, search is great in Google-world or on your Mac. But try it at work where everything is in disparate SharePoints if you're lucky.
> Hence, the idea of using manual labels to classify stuff - including emails - can still be useful. Crazy that we're forced back to this.
I know!
In my previous org we did this but had a couple of things which were hard to file:
- Where to put templates (letters, powerpoints to be used by more than one department/function). Is it marketing or communication or management?
- Should we put products or processes on top-level?
- How to split access control?
- How to archive (old/backup folders)?
One thing that stuck with me when I wanted to experiment with this system is the fear of exhausting the top level namespace: the system as described blocks the first number into areas of ten. I remember feeling like it would be pretty normal to need more subcategories in some areas than others and worrying about running out.
On reflection I think that's probably more of an academic problem, and that if I took the dive things would work out. I would be curious about whether there are any suggestions around for how to handle it if you're running short on areas to assign.
https://play.johnnydecimal.com/johnny.decimal/projects
This is actually what I’m running at work now. Over Xmas when I get some time an edited/updated version of this will become a page on the main site, it’s easily the #1 question I get by email.
The officer in charge of operations and plans for a warship can sit in an office with "N3/N5" on the door and everybody in NATO will understand what's going on.
But then they are unsuccessfully applied by people with a completely different, chaotic temperament, which are not able to maintain any such system for a significant amount of time anyway. We (as I'm certainly from the second kind) try one system after another, hoping that this will be the one that finally saves us from ourselves, but the real difference lies not in the system, but in something very fundamental in our psyche. It doesn't matter how beautiful the system is, or how fancy are the pens that we buy each september (school memories are the best example of this) - a couple of months later it will lie abandoned and forgotten.
May be I'm wrong? Has there been a single HN user that has been unorganized and chaotic most of his life, but then have actually turned things around because of such a system? Would love to hear about it.
But! Every now and then, something from one of these systems will stick for longer. Perhaps a technique, or vocabulary, or a piece of tooling. Over time, it coalesces into this chaotic, very idiosyncratic organizational system that kind of works for you. I'm definitely much more productive today than I was those two decades ago.
Those of us that are organised just need to help those people. Or be the person in our organisation that takes control of this stuff and does it. Hey, I’m making a decent living partially because of this system.
Animal asked you to think of an animal, then it tried to identify it by asking you yes/no questions. Let's say it asked you if it has four legs (no), does it have wings (yes), and then it guesses that you are thinking of a hummingbird.
That's wrong, because you were thinking of a penguin. It would then ask you what yes/no question it could have asked to tell a penguin from a hummingbird, and you could tell it to ask "Does it fly?" and the answer is "no" for a penguin.
It would add that question to its decision tree, so if anyone else thought of a penguin it would get it.
For much of my stuff at work, I use an unfiled filing system. I started this one day when I was asked to write some one time report to solve some mystery.
I spent a half hour with the line "mkdir " typed into my terminal window unable to think of a good name for the directory to do this task in. Finally I gave up on naming the thing, and did "mkdir unfiled". Then in unfiled I did "mkdir 1", and "vi directory.txt". In "directory.txt" I added something like this:
"1 Report for Alice and Bob to identify source of unexpected load on main website"
and then went into the "1" directory and wrote the code to gather the data and analyze it and produce that report.
I intended to figure out a real name for the thing later and rename/move it.
I never got around to figuring out a real name, and unfiled/1 is still there. Along with unfiled/2, unfiled/3, ..., unfiled/314.
This has actually worked out well for me. Suppose I run into some kind of issue with UTF-8 handling in MySQL. I remember I've dealt with that before, and grep in directory.txt for terms like mysql and utf. That turns up that the work for those earlier problems is in unfiled/258, and so that's where I go to find the programs I wrote for investigating that kind of issue.
Not everything goes in unfiled, of course. A good rule of thumb is that if it is something my replacement would need to takeover if I got hit by a bus, it should not be in unfiled. It should be in a well named repo on my department's git server.
https://en.wikipedia.org/wiki/List_of_Dewey_Decimal_classes
You probably don't know much about Dewey, he's got a pretty "interesting" life story:
https://en.wikipedia.org/wiki/Melvil_Dewey
"At the June 2019 conference of the American Library Association the Council voted to remove Dewey's name from its top honor, the Melvil Dewey Medal; the resolution cited Dewey's history of racism, anti-Semitism, and sexual harassment. The resolution was passed overwhelmingly with no debate. The award was renamed the ALA Medal of Excellence at the Association's January 2020 conference"
Yep, a friend of mine named it in 'honour' of Dewey Decimal. Kinda genius, which I can say because that wasn't my idea. :-)
But yeah had no idea he was a terrible human, that's a shame.
> He resigned from the American Library Association due to allegations of sexual harassment, racism, and antisemitism in 1905.
Or, you know, don't encumber the brains of people with useless numbers, and use computers for what they are actually good are, at least better than people, e.g. use the search feature of your file browser for "payroll schedule".
It's a common issue with intranets. There's lots of content, but not enough traffic to figure out what you may be interested in. And even if there was, you may not be after the "most popular" item.
This is a real example from a past life, and really works when you're shouting across an office or talking down the phone.
The longer original, edited down for brevity, went something like:
"Hey Kristy, where can I find the payroll schedule?"
"Okay go to the Finance folder. No, the one in the G: drive. Yeah now go to Payroll ... oh sorry, it's Payroll 2020. Yeah no not the Payroll 2020 in the Payroll folder, the one at the root. Yeah. Now in there ... hang on let me have a look ... okay there's a 'Janice' folder and in there..." ... and so on, ad. absurdium.
And search systems today are still terrible. Different interfaces, searching different locations and content, with different ranking and relevance and operators. Much faster to just direct someone where something is and let their brain figure out the rest, because we're actually very good at that.
This means that every year we get to recategorize the stuff underneath it and we learn as we go. Most of the sub-folders have stayed the same year to year, but some are different. It lets people see what works and what doesn't, and it prevents the cruft of ages to accumulate. Sometimes work bleeds from one year into the next, but if you're starting something new, create it in this year's folder.
Having been the victim of lots of shared folder catastrophes in the past, this simple rule has really worked out well for my team so far.
This is my gotcha: how do you decide where a document go? The categories might be clearly split, but documents typically won't fall in exactly one category.
How do you manage temporary tags? That is: how do you operate on the documents? You might be having some documents for your "2020 tax return" in several categories of your system. And what now? Do you duplicate them in another category (tax return?)? Do you tag them? How do you keep track that you moved them? (In case they're real objects, like pieces of paper).
Symlinks!
> How do you manage temporary tags?
Symlinks again!
Basically you put everything in an incoming folder, maybe organized by year or month/year or even type. Then you symlink to all of the categories it would be interesting to you.
Your temporary tags would just all be a folder of symlinks, so you can kill the tag and retain your document.
Symlinks can work, and I've used them.
My advice to people is usually, though, that your brain will sort it out for you. And it really does.
The canonical example is "home insurance". Say you have a category of "home" and a category elsewhere of "insurance". Perfectly possible.
Which one does it go in? Well, you designed this system. Your brain came up with it. I've found that in nearly all cases, you just know.
I see where this may fall down in a shared environment (work), which is why I advocate (in one of the many unwritten posts on my as-yet-to-exist blog) for the role of a 'librarian' at work. We seem to think that 20 people can work on a project together and all just magically decide where everything goes. That has been shown again and again to be a fallacy that we believe search will help us with. My argument -- from experience -- is that it does not, and that we need to go back to a world where we organise things, and that people are responsible for this organisation.
That said, the only documents I need to keep in long-term storage are invoices and contracts, which I can just put into two directories. If I ever get too many of such official documents, I might look into a structure like this.
Exactly unlike actual documents. Where do I put the sales figures we use for marketing?
This is what basically kills the whole idea. It's what made people want to switch from folders to search, which this site claims doesn't work.
Yep, this can be an issue. See my sibling comment here:
so something like:
## Marketing
- ##.## Reports (as a timestamped pdf etc)Now I just name things verbosely and use Everything (https://www.voidtools.com/support/everything/).
Oh yeah: Yahoo directory vs. Google.
I've given up on organizing documents in hierarchies a long time ago. Specifically because many things actually belong in many places.
Search is the way to go.
[edit]: and by "Search" and mean "indexed search"
/[CO] My Company/[OR] Organizations/[GOOG] Google/
I then put these prefixes in front of my files:
COORGOOGINV-20201214 Some Invoice.pdf
The number of prefix characters should be consistent within each folder of course.
I might sometimes add other tags:
COORGOOGINV-20201214 Some Invoice [No 345750][E 87][Payed].pdf
I don't find tags useful for many of the same reasons that I invented this system.
What's the tag for the payroll schedule? Is it 'Payroll'? 'Finance'? Something else?
Who manages the tags? Can anyone add one?
Tags need managing just as much as -- if not more than -- a well-organised folder structure. Tags aren't just some free magical way to organise stuff ... otherwise I dare say they'd be more popular.
"What is the tag?" is a category error. It's "what are the tags?" Tags are not paths.
Is it Payroll? Yes. Is it Finance? Yes, if that's how your org does things. Is it other things besides? Yes. A "superseded" tag would likely be useful to many organisations.
The idea of tags as opposed to paths is that a leaf object can have lots of them.
Of course tags are not free. The question is, does the value they bring outweigh their costs? Like you, I think the answer is usually "no". People suck at tagging.
If tags are to be used, "who manages the tags" is a key question that needs to be explicitly asked and answered at the same time as, or before, deciding your categories.
However the same applies to managing the "chart of accounts" that Johnny Decimal requires. The structure that is best for the organisation almost certainly won't be optimal for any given staff member. So management is required.
These are both stewardship problems.
The Johnny Decimal website is silent on the subject of stewardship. Here I want to focus on two aspects: keeping things working, and managing succession.
Keeping things working: using a gardening metaphor, is planting and weeding, re-potting, trimming and pruning and harvesting, turning over the garden beds, and so on. Like gardening, this part of records stewardship is best carried out regularly and frequently.
Obsoleting things is a key part of this. Contra the poster above who says "just delete it"[1], often you have to keep no-longer-current versions of filed objects around for a long time, for some combination of reasons: the maintenance and repair manual for a machine tool model we stopped selling 20 years ago, minutes of board meetings from an equivalent time ago, commemorative photos, old logo artwork, client records (if, say, you're a dentist) for the life of the client, even if you haven't seen them for a decade, ...
Succession is about ensuring groups have the right access, new people belong to the right groups and are properly indoctrinated, and things are handed over when staff leave or change roles, and ex-staff have their access removed; and about revising the structure, ontology, and access rights so that they stay relevant as the organisation they serve changes.
Records management is somewhat like plumbing--good drains, no-one notices; with bad ones, everyone is miserable. And keeping the drains "good" requires work.
The problems that are now universal ("Acme Corp Contract", "Acme Corp Contract - final version", "Acme Corp Contract - final with client revisions", "Acme Corp Contract - senior management approval", "Acme Corp Contract Brian", etc.) have indeed come about because computer-illiterate and filing-illiterate office workers thought the FAT file system (and then NTFS and/or Sharepoint) was a magic, zero-effort substitute for having a system both for keeping records and for training/indoctrinating new staff to keep records properly.
We have had two generations of office workers who have never experienced properly functioning filing systems (except possibly at a university library). Just like people who browse the web without an ad blocker, they don't know any different, so this is an invisible problem to them.
Johnny Decimal is a step towards opening people's eyes (or noses). Thanks for publishing your website!
1. Deleting things that should be deleted is an important skill, but another one is knowing what should not be deleted, but instead moved out of the main stream.
An error I intend to correct! On sufficiently large projects I believe you need a ‘librarian’ who has ownership of the system. I’ll expand on this on the site eventually.
Ah, if only I didn’t have an actual job. :-/
> We have had two generations of office workers who have never experienced properly functioning filing systems
This sentiment inspired the opening paragraphs of the site. We literally used to have people - women - called ‘administrators’. Their job would be to file things so that people could find them again. Then we all got a computer, the number of files exploded 100x, and we pretended like we could just search for what we needed.
* over 5,000 entries * over 3,000 tags
And I still have issues finding posts I've known I made via the tags.
You can have additional tagging and full text search on top of it.
When I go through old documents, I don’t even skim through the contents. I just check the date and remove them if necessary.
A whole book on this is "Delete: The Virtue of Forgetting in the Digital Age" by Viktor Mayer-Schönberger (2011).
A longer piece with the author in The Guardian: https://www.theguardian.com/technology/2011/jun/30/remember-...
And the book: https://www.jstor.org/stable/j.ctt7t09g
Otherwise you end up with a digital ball of mud, and you can only extract info from your system using the search. Maybe this works for you, maybe not.
I like that Johnny.Decimal assigns memorable ids to the categories, this is absolutely nice and is easy to follow by members of a team.
This seems like an interesting system, provided you don't have more categories than you can easily remember.
If you use it in a team setting, there must not be more categories than the team as a whole can remember which is, perhaps counter-intuitively, fewer than you can probably remember on your own.
As with all classification systems, one should beware the Celestial Emporium of Benevolent Knowledge: https://en.wikipedia.org/wiki/Celestial_Emporium_of_Benevole... or be doomed to repeat it.
One particular modification is to extend the pipeline between the "find" and the "xargs" to filter the file list, especially `grep -v .git` ;-)
Search today is still terrible. Nothing reaches across all content or platforms easily, and different scoring systems and operators in different interfaces means it's just as big of a mess as the files it's meant to find.
The ability to direct someone else to a location is also incredibly important, especially because "where" something is can be just as important as "what" is in it. Most systems seem to forget or abstract that away, to their own detriment.
New Folder New Folder(2) New Folder(3) New Folder(4) New Folder(5) New Folder(6) New Folder(7)
Muscle memory can be important as well. If I'm going through the Finder/Explorer, being able to pick the 4th item down and having it be the one I'm concerned about is helpful.
Johnny Decimal also helps with avoiding having too wide of a hierarchy, as well as having too deep of a hierarchy. Can't be more than 10 folders wide at the area or category levels, and can't be more than 4 levels deep.
People do just naturally put numbers in front of things, though, and I’m convinced it’s to solve the muscle-memory issue. Or it’s that I generally work with nerds and we find it easier to remember that our folder is 34 in the list of 67 rather than remembering that it’s ‘SQL designs’ or whatever.
I mean literally back at a major bank 15 years ago the root structure just had numbers in front of every folder. Long before I meddled with it.
e.g. Accounts Receivable #15000, A/R for Foo Corp #15001 and so on.
Thanks! I put a lot of work in to it and I'm no designer let me tell you.
FWIW it's running on Gatsby. SHOCK HORROR JAVASCRIPT. ;-)
Helps a lot that it's very light on images.
It’s a real PITA to be honest. I’m typing this from bed and can’t remember the exact details but chuck me an email and I’ll have a look over the weekend. (Or just scan the repo if you can be bothered, but I’m happy to try to remember what I did and put it in a mail.)
https://github.com/johnnydecimal/johnnydecimal.com or hello <at> the domain.
The site is written in MDX and then `gatsby-plugin-mdx` does the heavy lifting. What I could not get to work was Markdown formatting for inline images, especially not for smaller images. They always ended up being shown at 2x.
So now it's just an image, imported in to the file at the top, and displayed further down.
import outlook_folder_structure_1232_expanded from "./Outlook__folder-structure-12-32-expanded-min.png"
As for making them sharp and small, they're just screenshots from my Mac run through one of the many online PNG resizers. From memory that gave me a file with a 78% saving and if it wasn't near magic then I couldn't tell you how it did it because there was zero difference to my eye. I can't remember which one I used, just try a few.
What I can say is that I like it for helping you choose levels of hierarchy. So far I've squeezed 10 or fewer sub-categories into each chosen category. (I use the 0s, so within 10-19, I start with a sub-folder of 10...)
What I do not do is any of the other suggestions - searching by it, memorizing numbers (though I have an intuitive sense now when organizing email or looking through it.)
I have not tried this folder structure yet for files.
I really like the idea of 'everything related to one task in one place'. What I'm planning on doing with the app I'm (trying to) write is extend this to make sense of the fact that your task is in a JD category and might therefore be related to other tasks in the same category.
So I'm looking at `11.01 Some item` and it has a bunch of tasks. Over in `11.04 Another item` I have other tasks, and optionally I can choose to show them -- because they're likely to be at least related to the thing I'm doing. Maybe I can do a bunch of related stuff at the same time?
One of the things I love about JD is the way it helps me stay in context. Everything in a category relates in some way to each other, so given that my job is mental and I do a hundred things a day, I can save myself some cognitive load and at least bunch similar things together.
I see a lot of comments claiming that either tags are superior or this is bad because you need an inventory for your system. Both ignore a very real problem: with tags you'll end up with near-identical tags which doesn't really improve your ability to find stuff. You can have a hierarchy of tags, sure, but then once again you need some sort of inventory of tags.
Which brings me to my observation: if you think that tree hierarchy is limiting you and you need free form tags, your hierarchy is probably bad and you're simply tagging things to feel like you're organized.
> You never will. And if you do, the category you defined was almost certainly too broad. Split it up.[0]
Alternatively, your IDs are too narrow, so you should be stuffing more into each ID folder. That is, to modify an example from the website: You have "14.01 1601 Timesheet". But you have a new ID for every pay period. So you have "14.02 1602 Timesheet".
Better would just be have "14.01 Timesheets", and each timesheet gets the year and week prepended to it.
He discusses this a bit in the exceptions page[1]. This is where he both makes subfolders, and talks about storing things with dates in them.
[0]:https://johnnydecimal.com/concepts/ids/ [1]:https://johnnydecimal.com/concepts/exceptions-to-the-rules/
This is what I do. Isn’t the site clear?
I (still, 5 years later) have `14.01 Timesheets`, and within there I have `yymm` folders which contain each entry.
14.02 is still free. So yep, a ‘wasted’ category there ... but I have a hundred to play with, and really rarely use more than about 30 in a complete system.
The Internet Advertising Bureau uses a similar structure to categorise things. For example:
- IAB20 Travel
- IAB20-1 Adventure Travel
- IAB20-2 Africa
The problem is that sometimes you cannot find an appropriate category despite there being hundreds of them defined.
! cdj script is not found on github https://gist.github.com/jen729w/cab3a02bce70dab7a4567d460d5c...
"In payroll (pause) schedule" (which you'd interpret as /payroll/schedule)
* 10/11 notes/
* 10/11 notes/11.01 notes/ (... all my text notes)
+ (I usually use a shortcut / alias of `notes` since it is the most frequently accessed directory)
* 10/11 notes/11.02 onenote/ (... all OneNote docs)
* 10/12 home/
* 10/12 home/12.01 123 MainSt/ (... deed, map,...)
* 10/12 home/12.02 insurance/
* 10/12 home/12.03 HOA payments/
* 10/13 finance/
* 10/13 finance/13.01 paychecks/
* 10/14 tax/
* 10/14 tax/2020/ (... tax by years, this is one place I use year instead of sequential id#)
* ...
* 20-29/21 projectX/ (... current working project)
* 20-29/21 projectX/21.01 src/ (.... this is usually from git)
* 20-29/21 projectX/21.02 assets/
* 20-29/21 projectX/21.03 docs/
* 20-29/22 projectY/ (... side or 2nd project or project I'm planning on working in future )
* 20-29/23 blog/ (... all my blog post source)
* 20-29/23 blog/23.01 siteX/ (... hugo blog post files for siteX)
* ...
* 30-39/31 JSFramework/
* 30-39/32 VideoResource/
* 30-39/33 AudioResource
* 30-39/34 Reference/ (...programming books, doc related to work)
* 30-39/34 Reference/34.01 javascript
* 30-39/34 Reference/34.02 reactjs
* 30-39/34 Reference/34.03 dotnet
* ...
* 40-49/41 tv_movies/
* 40-49/41 tv_movies/41.01 futurama/
* 40-49/41 tv_movies/41.02 trailers/
* 40-49/42 music/
* 40-49/43 os/
* 40-49/43 os/43.01 debian/
* 40-49/43 os/43.02 ubuntu/
* 40-49/43 os/43.03 freebsd/
* 40-49/43 os/43.04 windows/
* 40-49/43 os/43.05 mac/
* 40-49/44 games/
* 40-49/44 games/44.01 DOS/
* 40-49/44 games/44.02 retro/
* 40-49/44 games/44.03 mac/
* 40-49/45 books/ (...books not related to work, ie fiction, non-fiction)
* ...
* 50-59/51 archive/
* 50-59/52 backup1/
* 50-59/52 backup2/
* ...
I can use this to partition type of files I have into separate external disk or flash drives, too. For work, when I'm on the road, I can just have the folder 20/* and 30/* in an external USB drive instead of taking the entire NAS. If I need to escape a burning building, I can just take the small SD card which contains 10/* backup.
It also helps me with backup strategy, as 10/ gets backed up most frequently (ie daily/weekly), while 40/ gets backed up only once in a while.
(EDITED to show example directory)
In most situations — you’ve found an exception — it’s better to think about it a bit but then just start. The numbers have no meaning, even if that doesn’t tickle my neatness gene quite as much as it might. ;-)
Yep, and the numbers tie things together across disparate systems, e.g. your file system and your to-do system.
So I have a project, `12.34 Whatever`, and it has some files in the file system. I also have an OmniFocus project named the same. And an email folder.
It's real easy to line those things up together when you want to work on that thing.
This scheme is the best I’ve found (some tagging software used it): `report[important,sales,veronica].pdf`.
>Now those same people store all of their files in arbitrarily named folders on their company’s shared drive and wonder why they can’t find anything.
This is the most outrageously false premise on which to base an argument. It's staggering.