Johnny•Decimal – A system to organise projects
This website surfaced in my RSS reader. It's something of a tagging system, with ideas possible to implement both for a folder structure but also, I think, to organise a tag structure in a Zettelkasten.
One of the benefits I can see in applying this, or something similar, is that the same way to organise things can be used in many different areas: Folders and files, email, Zettelkasten, bookmark collections etc.
I haven't tried to impemented myself just yet. Feels like it's better to let the ideas percolate for a while before deciding on areas and catogories.
Anyone who categorise things like this and can share their expericence?
(For more complex needs, I also found this from the author's Twitter feed.)
Howdy, Stranger!
Comments
@thoresson Interesting web site. It teaches one concept that I figured out some time ago on my own, which is that you don't want a folder system on your computer to have too many levels, otherwise, you can't find anything. I stick with 2 levels (preferably); on the rare occasion 3 levels. But I like the idea of using a "decimal" system to label everything.
I'm not sure how it could be applied to a ZK. It might be applicable to tags; I'll have to think about it. But I wouldn't want to use it for zettels - the structure is too hierarchical.
You might like to look at Brett Terpstra's post about punctuated tags https://brettterpstra.com/2017/08/09/regarding-punctuated-tags. He is the ultimate geek and often has interesting things to say -- and useful utilities that he gives away.
And incidentally, he is another advocate of the shallow file system (as he states in this post) https://brettterpstra.com/2013/12/20/automatic-filing-with-hazel-and-mavericks-tags/.
the method described in the first link immediately reminds of the A-Z alphabetical filing system in the book "Getting things done". Instead of numbers it uses letters and it is only one level deep. It shares the same benefits from the ones mentioned in the linked article.
I like how the author in your linked article provides more insight into ups and downs of the method, and how he is trying to solve some problems.
here is an interesting discussion about advantages and disadvantages of categories.
It shows that categories need to set exclusions (where absolutely not to search in) but also taxonomy (synonyms, antonyms, etc).
I suppose the first level plays a mayor role in defining exclusions so they must be chosen well. Simultaneously it must be arbitrary enough to avoid ambiguity. Depending on your individual needs this can be easy or challenging.
Is it possible to create a better structure?
I can not refer to any documented method, but i've experienced that the reason why these schemes work so well is because you create a mental copy of this structure very fast. I've seen this when guiding a friend 7 levels deep through the windows registry while being outside on the phone. Or when i click my way through
C:\Users\chris\AppData\.... What i mean by that is that the brain behaves incredibly responsive toTwelve dot oh-threein a short learning time. Unfortunately i haven't made the change to a 2 level depth system for my projects, yet.my first Zettel uid: 202008120915
Thanks for sharing this. My two cents:
¢: I don't see this system necessarily working well for a ZK without having a Zettel as a tag index. But, I do think the principle here, applied to tags, makes good sense. (One example from my own practice: I only have the tags
#GISand#cartographyfor notes related to those areas. Many notes have both, some only have one or the other -- but I avoid the specificity of tags beyond that, because it gets beyond the "two-layers deep" idea).¢: On the other hand, I find tagging other files on my machine to be a pain, so some sort of clear semantic structure like this outside of a ZK seems right. If the ZK
Notesfolder and the folder of academic references exists outside the AC.ID system, then there would probably be less clutter when focused on knowledge work, too.Is anyone going to try this and see what happens?
Whenever I look at the catalogue entries of the University Library here in Cambridge, I am usually presented with quite a collection of keywords for each item. Herein lies the problem, and one that I have been wrestling with for years: categorisation is difficult. It was OK for Linnaeus and his plants in his garden (not to belittle his achievement, the effects of which we are still feeling, and which will continue to echo down the ages), but it is a bit different for me. What am I to do with a note that equally fits the categories of psychology (social), psychodynamic psychotherapy, history, biography, politics, and philosophy? Perhaps is it my fault for having wide-ranging interests (or being indecisive, as some would have it) but that is what I am faced with on a regular basis.
Hence, I find that I am slowly (and indecisively) moving away from categories and towards themes (which some of you will say are just another kind of category, and you may well be right). But to me there is a psychological difference, and it is leading me away from putting things in folders in hierarchies. Hierarchies are damned inflexible things, and they always seem to cause me trouble. Associations are perhaps more my thing these days.
That is why I like tags - you can have a few or more and you can do a Boolean search on tags. Combine that with an "organic" network created when you link zettels and your ZK is organized in a much more flexible and accessible manner.
@GeoEng51 , I like tags for the same reason, in my ZK. Nevertheless, I'm not sure even the creator of this folder system would use it for things like a Zettelkasten. He states that he keeps a folder
0for a place to store a folder that's beyond those folders, or the ground out of which other things grow. Couldn't something like a ZK exist there?To the same point, to what extent do others like outside files in their ZK?
(Note that I'm not defending this system, only a curious experimenter who is considering it for everything else outside my folders for notes and academic articles)
I always love to read about systems like this that have been tried to organize paperwork efficiently.
The best thing that ever stuck with me was 1 Filed Documents folder, and in there folders for each year, and in these I put all the scanned documents and stuff to keep that I received during that year except two folders where the items are strictly sorted by date:
That's how far David Sparks "Paperless" ebook got me back in the day
Without a full-text search, I'd be in trouble, though.
My project files live in a separate place, though, and from @thoresson's 2nd link to https://play.johnnydecimal.com/johnny.decimal/projects I gather that the author found up to 10 client subprojects to be lacking in the long term. You don't have the problem with data-stamped free-form folders in a "Work in Progress" directory. But then you have to clean up from time to time to keep the WIP dir small enough to make sense.
If anyone here has a sophisticated system for filing, I'd very much like to encourage you to open a discussion and show us your stuff
Author at Zettelkasten.de • https://christiantietze.de/
The nub of the question, it seems to me, is this: how do I find something when I need it?
If I use a folder hierarchy I am faced with a number of decisions, both at the time of filing something away, and at the time of searching for it. The first question is "where do I put it?" And given that much of what I have might reasonably fit into several different categories, this is not always simple to answer. At the time of seeking/finding, the question is "where did I put it?" And once again, I am faced with the same question regarding categories. So I have to remember or guess what category I used for the item.
Alternatively, I can use full text search (for example, HoudahSpot or DEVONthink) and then I'm not sure it matters what folder the item was put in. Unless one wants to narrow down the search area, but I'm not sure that saves a lot of time.
It seems to me that folders are a hangover from the days of paper when if you wanted to find something again, it had to be filed in a particular place. Now, folder hierarchies seem less and less useful to me as time goes by. Indeed, I begin to feel that they now take up my time instead of saving it. Deciding where to put things and then curating the contents of folders to make sure things have not got into the wrong place is a drain I can do without.
I agree with this, I'm setting up a system where I access/find my notes by using search (with GREP, which I can use on my laptop and Android phone). All my notes will be stored in one (synced with OneDrive so I can access it with multiple devices) folder.
In fact, as I was standing in the shower this morning (why do ideas always come at that time?) it occurred to me that I could have a "Zettelkasten" of sorts for almost anything -- bills and receipts, software licences, you name it. OK, it would not really be a Zettelkasten, but I could have the flat file structure, associate items with tags if necessary, and have unique file names (if necessary). And I should still be able to find things. Only the most basic filing would be necessary. Heck, I could probably have one folder called Home and another called Work and still find anything I needed. (Though I believe some aspects of performance are affected by having large numbers of items in a single folder. Experts would have to comment on that. Would it affect the speed of full text search?)
Hmm I do use timestamp Ids to the second (20201217170123) for recipes/invoices and other filing things with success. Combines the ubiquity of full-text search, so place is not important, with the unambiguity of direct links -- unlike a search for "invoice hard drive", I do not have to pick the correct one I'm looking for from all the matches, as there's supposedly only 1 relevant match.
In terms of overwhelming the OS -- I find that macOS Finder sometimes has to refresh for a noticeable bit when I look at my ~6000 note folder. But it performs well. Windows is known to have trouble with access speed to large groups of very small files (as used for source code in large projects using programming languages for the web), but y'all can check out our 10000 Markdown files download and duplicate the files a couple times and see what happens with 50000 files: https://github.com/Zettelkasten-Method/10000-markdown-files
This is a nerdy solution to the filing problem. So I assume people interested in trying this are eventually competent enough to whip up solutions. There is a multitude of command line applications that are blazing fast when it comes to search in a folder full of text files. PDFs etc. are a different beast, but there's solutions to index these files as well, no doubt. I don't know if they are as user friendly as Cmd+Space for Spotlight on macOS, for example, though. But I am convinced the situation can only get better, not worse, in that department: it's easy to whip up a good-looking graphical user interface for a good command line application and provide some eye candy.
Author at Zettelkasten.de • https://christiantietze.de/
This isn't directly related to the original post by @thoresson but I thought an anecdote with some relevance would be entertaining.
I used to do a lot of geotechnical engineering work for the two main railway companies in Canada (CNR and CPR). I quickly learned that they had an amazing system for finding information, no matter when it was produced. Their system was manual. They essentially had a very long filing cabinet that was organized linearly by distance along each of their rail lines. So, if I wanted to find something related to Mile 92.34 on their main line east of Revelstoke, I just went to that location in their linear filing cabinet. Then, at each location, the information (it might take up a few inches or a few feet, or more) was organized by time - earliest to the left and latest to the right. Some of those files started in the late 1800's.
This was a very efficient system and as long as people were diligent about their re-filing of information, almost foolproof. It was also very informative for the casual browser.
Along came the 1990's and computers, and those wonderful paper files were discontinued in favour of various computer-based databases (a number were tried). The result was diaster - no one could find anything.
I'm sure the situation has improved since the last time I checked, but there is a good lesson there for us. On occasion, paper files are extremely efficient and useful; we need to think carefully before we replace them with a less efficient, less useful electronic alternative.
@GeoEng51 thank you for sharing this. It is interesting that this system was essentially 2 levels deep. I find it somehow depressing that i, in search of a new system to organise my files, discover the 2 levels rule as something new when it actually has been used way before 1990s.
Do you have any thoughts why the new system was a disaster? The manual system could have been continued on the computer's file system. Software design was not intuitive at that time, i've heard, but my earliest childhood memories with Computers don't reach this far in the past.
Have these complains reached the IT department? Did they ever returned back to the manual system or was it too late to turn back?
my first Zettel uid: 202008120915
@zk_1000 I believe the main issue was that there was no way of easily converting all the paper material (all of which was on various sizes of paper, not just the drawings but also the letters and memos) into electronic form. It wasn't like today, when it is a snap to scan and store documents in PDF format (PDF didn't even arrive on the scene until 1992, and then only in a rudimentary form with expensive software). As I remember it, the "brave new world" of the paperless office took a good 10 (if not 15) years to materialize in any practical way.
So - it would have made sense to keep the paper files in their linear filing system, and just build an electronic database around them, so that you could search for the material on your computer and not have to look at the physical files every time. But you'd still have the physical files and the "old" system as a back-up. However, in the wisdom of the IT people of the time, the linear files were organized in some other way (I'm not sure what; hopefully not totally randomly, but that might have been the case) and then whatever database software was employed (also rudimentary and expensive compared to today's offerings) did not work very well. Worst of both worlds - you now can't find the information in the computer and you also can't find it in the paper files.
I suspect if one took on a problem like this today, with better tools and more experience with large databases, the conversion would occur much more smoothly, effectively and logically.
I'm sure there were a few IT people back then who were tarred and feathered, or close to it. Unfortunately, I don't know the end of the story, except to say that CPR and CNR are both in business today and the engineers continue to be able to do their jobs, so I guess all was sorted out eventually.
It was a real pleasure to use the old linear, paper system, though. I try not to be nostalgic about the old way of doing things, but some of them had a sort of attractiveness and beauty to them that you don't see today. I recall reading through an early 1900's report on the design of a railway tunnel - the engineer in charge was a true artist and had taken an afternoon or two (doubtless in pleasant weather) to sketch the hill through which the tunnel would traverse. He then added his concept of the tunnel into the sketch, without actually having designed anything yet. Mind you - he wasn't wasting time or procrastinating, at all. It seemed the first stage of conceptualizing was to create an object of beauty (at least, to the eye of an engineer, but I think also to some extent for the general public). Those were days when the marriage of form and function was the responsibility of one or two people, and they'd darn well better get it right! The same applied to the design of buildings and bridges from that era.
Hey, Johnny-as-in-Decimal here. One of the forum members here told me about this post. Just here to say hi and offer help/to answer any questions/to look around and get ideas. I love these various organisation systems, although a tag-based one isn't for me personally. (No judgement at all, we're all different brains and that's wonderful.)
I found your article, referenced by @thoresson, was quite interesting; thanks for sharing those concepts.
I just wanted to point out that Zettelkasten is mostly not about tagging. Tagging is a small tool that can be used, but the main concept is one of connecting notes in an organic network, suited to how each person thinks and evolves thought patterns. There are hierarchical tools as well, such as "structure notes" and tags, but the meat of the technique has nothing to do with hierarchy. Stick around and read through the Getting Started material; you might find something of interest.
I'm glad someone recognises that! It is not as common as one might think ... see https://en.wikipedia.org/wiki/False_consensus_effect.
But it's not just different brains -- it is different fields (of work and study) and different material, as well. I habitually work with material (in psychology and history) that simply does not fit in hierarchies. The richness of the material means that it usually crosses any boundary that you can think of, and merrily bleeds into all sorts of areas that you might not expect. And to me, it is where material (and study) crosses boundaries that things get interesting. So I can use hierarchies to some extent, but it is crucial to me to be able to use systems that break out of hierarchies and make connections or associations that go beyond simple categorisations. The human mind is by its very nature associative, and making associations is what matters most to me in my work. Multiple associations are even better.