Folders
Hello.
I'm aware that it is generally good practice it get rid of folders, but to what extent? Can I have a folder for each article? What about when I'm reading a book or textbook and it is organized by chapters? Can I create folders for Each chapter?
Any help would be appreciated. Thanks.
Howdy, Stranger!
Comments
Welcome, @StellarThread !
I'm going to respond to your questions within some questions. Hopefully, it will help us to discuss and think through some of this more.
You said:
Could you explain why you think it is a good practice to eliminate folders?
I think that discussing some of the underlying assumptions first will help us answer some of the following questions.
Your questions are pretty vague (not clear what your goals are), but I have never let a lack of background and insight prevent me from offering an opinion.
If your notes and files are searchable by keyword, (or if paper notes you have either a searchable keyword index text file or alphabetical keyword index cards) there is no need for folders for each reference or topic.
I use two folders on my laptop for my ZK-related stuff. 1) Inbox and 2) Sources.
"Inbox" is articles to read, and "Sources" is to store those that I have read.
My notes are on paper index cards. I use a text file for an alphabetized keyword index.
Spend time thinking about organizing your information by searchable concept rather than putting everything in separate folders.
Hi,
it's not easy to give an universal answer, and there would be a lot to say.
Use of folders, just like tags and links, has benefits and drawbacks and its impacts depends on context of use, but also our personal attitudes. Folders are a rigid construct, and this can be a thing that I can feel comfortable or uncomfortable with.
It depends on the software you use, too.
Folders are useful when you need to keep a bunch of things close each other and separate from other things.
Ask yourself, do you need to take notes born from a book physically close to each other and separated from notes born from another book?
In my case, for example, is not useful, on the contrary is an obstacle for my way of developing knowledge. I need to "merge" knowledge, contents and informations about the same thing taken from different sources.
Do you need to move or transfer an entire set of notes from a place to a new place? Using folders makes this task easier, because a folder is a physical natural container that makes a set of notes one thing.
Do you need to put the same note in two sets? In this case use of folders is a bad idea, a note can be only one parent folder.
You can ask these questions at different zoom levels in your set of notes (chapter level, book level, subject level and so on), so you can find how wide they should be your ideal folders.
I can tell my experience. Talking about Zettelkasten, during time my approach was gradually abandon their use. When needed, I obtain the required order and organization linking my notes in a "structure note", rather than put a set of notes in different folders.
Linking notes in a structure note rather than organizing them in folders it's a more flexible approach.
In my case I use a single folder for an entire subject (photography, developing, knowledge management, and so on), but rather for legacy habit than real necessity.
This for the zettelkasten part of my note system. In my system I've "projects", too, and for projects I feel useful having a folder for every project. Every project need to be separated from others.
One thing to consider, anyway, If you don't use folders and don't use another system for structuring your set of notes, you'll inevitably obtain very much disorder. Disorder could be a good thing (helps to develop creativity), but also create problems when you have to reuse notes later. If you need order at some zoom level and you don't use folder you need to use another thing.
So, for example, you can put your notes about a book in a single folder, but if you need to maintain the chapter-order origin of your notes, you need to add another structure that replace the function of chapter folder for that. A structure note that outlines your book notes is a good fit.
I don't know if this fast writing could be useful.
If you tell more about your needs, we can better develop the solution around your needs
Right. Sorry, I was going to add more info but forgot to come back:
For one of my projects, I am reading a textbook on meta-ethics which explains various competing theories on the subject. I had a strong urge to create a folder for every author/theory within the book, but I remembered the advice from online going against that. Creating a folder seems harmless in this context, but I want to be sure.
Right, so I think we may or may not be on the same page. I have a file "contemporary meta-ethics" which has lots of concepts that are searchable. "contemporary meta-ethics" is also searchable and is tagged. I also have lots of other files which are searchable and not in a folder. But adding more "files" within the parent file "contemporary meta-ethics" would turn the parent file into a "folder" as I understand the term within my software.
Interesting. I might try that. Since I started I do have a lot of articles in my "literature notes" folder". I organized it by permanent, literature, and fleeting.
Right. That is where I am at right now. Still new to it though.
For sure!
If you use tags instead of folders, you can narrow down your search for any project you have or idea you are trying to find much more quickly. For example, for a university group reading I read content related to utopia, and so all of the articles I was interested in for that particular topic got tagged with utopia. A couple of the articles touched on population ethics, though, and so I tagged those articles with population ethics as well. As I continue on and do other projects I might read more articles related to population ethics as well, and so if I ever want to explore the ideas in population ethics a bit more to create a permanent note, I can get a list of all the articles/notes I have from the various authors and consolidate the concepts.
When it comes to my example, I have a book on meta-ethics that explains various theories, and so I wanted to create a folder for each theory within the parent folder "introduction to contemporary meta-ethics" which a friend recommended.
Do you generally agree with that approach? Are there any corrections or ideas you would want to add?
Interesting take. Perhaps for stem subjects like math this is useful, but I wonder whether this methodology would be good for authors who have competing interpretations? For something like philosophy, even for authors within the same branch and with a lot of the same opinions, they have different perspectives and their arguments differ in subtle ways. Hence it is better to keep them in separate folders.
Right. In my case I have a software in which I can use tags, though, and so there is less need for a folder.
Well, in my software, if I have a note in one folder, then I can create a portal to it in another folder, editing that note from the other folder. I get what you are saying though. The same thing can be done with tags. You might have a different way of doing it.
What do you mean by structure note, and how is that different from folders?
Yeah I had that same approach for a long time. Recently I am trying to experiment with using tags for each project, because projects might overlap. With a file for one project, it might also be useful for another project, so it is useful to have the extra tags there to help extract it for whatever area it is.
Right. There is a very strong hierarchy for my school notes, organized by semester/class/lecture/chapter. I can find things very quickly, but it does not feel conducive to creativity for sure.
>
Sorry, I am at a loss for what structure note means here again.
It was good! It is good to get everyone's perspective on it. Thank you.
Ok. Now it's clearer.
Just in this paragraph, I personally see a drawback on using folders.
You have two dimensions of structuring, by "author" and by "theory".
But using folder, you are forced to choose author as primary and theory as secondary (or viceversa). And what if in the future you need to reorganize the structure using the other as the primary?
What I would do (but this doesn't mean is the RIGHT method).
N.B. I don't know anything about meta-ethics, so I'll pretend, basing on the table of content of this book: https://www.routledge.com/Metaethics/Shafer-Landau/p/book/9780415439206
I do one single folder, Meta-ethics. I put all my notes about the meta-etichs book in that folder.
I do the main Meta-ethics note.
I start, and I create the first "theory note".
I don't know, the first pages I meet could be about "The Subjectivity of Values" (is this a theory?) of "L. Mackie" (is this the author?).
So, I create three notes
In the last note I write what Mackie states about Subjectivity of value
In the first and second note I put a link on the third note, and a brief intro about the author and the general theory.
If there is anoher position on subjectivity value, by John Doe, I create John Doe note and repeat
If in the book there is a second theory from Mackie, I create another general theory note and another instance theory note with the ideas of Mackie about that theory.
So, you have these notes:
...
When in Theory1 note you have a set of many author's theory instances, you can start comparing the instances into the note, developing in that place.
Surely Theory1 will have a "Macrotheory1" as parent. If you need, you can create that parent note, too. You can move up until you reach the main meta-ethics note.
Theory 1 can relate to Theory38 in some ways, so you can link toghether.
It seems awkard, but this model has big advantages.
It's not important the details of the model (creating authors notes, rather than general theory note, instance theory note), but the principle of moving the complexity of the structure from use of folders to use of notes and links.
You "lose" the concept of chapter, but in this case the chapter is not useful. The chapter is a physical structure of your source, not a structure of the concepts.
A very big advantage of this model make the things clearer, maybe. If you read a second book, or an article about meta-theory in the future, if you have this structure you don't need another empty structure for the second book. You integrate new concepts from the second book in the same structure you'have already made.
When in the future you need to revisit what you have about Subjectivity of Value, your entry point is Subjectivity of Value of note and moving between notes using links you've made you easily traverse again the knowledge.
The strategic thing to do is making "right links" between notes. That's all.
It's a very raw description, but I can't do better withouth a minimal knowledge about your subject
In very simple and inaccurate words, a "structure note" is a note that has, as its content, mainly links to other notes. The simplest implementation is making a bullet list of links.
If you have a set of notes, for example theories of Mackie, you can organize them in three way:
this is a very basic example of the Mackie note that is also a structure note:
it has the same function of a folder that contains the three notes. But it also have other big advantages.
This is a structure note that models your original source book, if you need this structure:
As you can see, you can create as many structures as you need with the same theory notes, while you can put a note in a single folder only.
.
Okay, I see what you're saying. I actually had a skeleton of notes for what (I think) you described before even chatting here. Here is what I have so far:
Is this what you had in mind? So if I read an article about cornel realism for example I can add another author that explains such a theory. I would agree that this model is more flexible. I added a bit extra, to show what it would look like if the author introduces an unrelated "theory X," which perhaps relates cornel realism to psychology or something.
Would you say these images demonstrate an understanding of what you said? I am really unsure of whether I understood you correctly, but the images above show the extent to which I understood you.
As I understand it, this would fall under the "put all of these notes in Mackie folder" you mentioned earlier, rather than the other two bullets.
If I wanted to try the tag method, I would do something like this for Nicholas Sturgeon:
I also added an alternative method which is basically equivalent, where I reference the Nicholas sturgeon concept. Both ways give you a way to get back to the original concept, as you can see where the reference is shown in the the top right for the author:
So far, I think I might either go with the reference method or the tag method. I am not sure which. Which method do you think would be better in the long term?
One problem I see cropping up is having to write a reference or tag every time I explain a theory, when I could just have a folder, however, I am guessing that the reference or tag method would have better longer term value, since an author can pop up in multiple areas (for example, William James as both a philosopher and psychologist. You would have to have separate folders for William James in psychology and philosophy if you use the folder method like in the first image of this comment)
Is everything above correct/valid? Do you have any corrections or things you would add?
As for the structure notes, I do not see myself using them, since I can create search portals within the folder for Nicholas Sturgeon which filters by topics like meta-ethics, epistemology, and the like.
I have some difficulties understanding the tool you are using. What is its name?
Does the screenshot you have posted an image of a single note?
If I understand well, items with icons are folders. things like natural properties are links.
In your software (it is an outliner, maybe) it seems you can make a note that contains folders.
It seems your software use a very different model from mine.
It seems that in your approach you obtain a long note about meta-ethics and put all in a single note.
Having a structure like this is a way of structuring things. It is very different from the way I model my things (rather than having one single note in which I put all the stuff about 20 concepts, I have 20 concepts notes and the main note contains 20 links to concepts notes and some sentences around links at most).
But if your software works very differently from mine (I use Obsidian) it could be difficoult to apply the same model.
I can't say if this can work for you. It could be, why not.
it's a matter of trying for a while, see if it works. If it works, maintain the method, if it doesn't work, change.
"works" mean: if you do this stuff, you learn well meta-ethics concepts, it will easy to find these concepts into your note in the future, if you can reuse these concepts in the future if you need them for doing something (writing someting, explain them to someone, and so on).
This is RemNote.
It does have a very different model. I use it for my spaced repeition memory system. It's model is very different from Obsidian which I think you are using or the Archive. They call every bullet a Rem, and they say, "Everything is a Rem." So, while they have "folders" and "documents" they are really just Rem like anything else, they can just do some other unique behaviors feature-wise.
And, @StellarThread is trying to use their spaced repetition system via the up and down arrows next to Cognitivism, etc.
RemNote doesn't lead itself towards: "Hey. Here is one folder that you put a ton of notes that are independent containers that you can link and structure however you want." The Archive does this naturally and effectively.
Here is what I would recommend with RemNote (and make sure to adapt this over time for yourself as there is no one coming down from Sinai with the 10 note taking commandments for perfect knowledge work here):
For your metaethics text, you create a "document" rem for it in the "knowledge resources" folder. When you read and takes notes on that text, you do that within the document rem for the text. For example:
All your coursework/project work notes (for say like lectures) go into the current projects/sources "folder." You just organize your work there. However, whenever you are coming to the end fo the project/course, you'll move things to the "ZK" folder.
I recommend just accepting the ZK as having some main topics (e.g., I have Philosophy and Epistemology as a child node). Set up things topically in a manner that seems good and aligned with your research interests. You can change things up quickly later if needed. Then you just move it to where it seems most fitting in the ZK. It's somewhat of a Folgezettel set up. Since you can use the portal or rem reference feature, it doesn't matter where you put it ultimately.
If you want to concept/topic in the ZK "folder" link to a key text in the "knowledge resources" "folder", you can with the "link source" feature. For example, I have "Popper, K. R. (2002). Conjectures and Refutations: The Growth of Scientific Knowledge. Psychology Press." as a linked resource to "Epistemology" as for me, that is one of the most important texts on Epistemology for me right now. That might change later.
Again, there is no right or wrong. And the software you use will encourage different behaviors.
Using something like the Archive in many ways is much more simple. But, you loose out on the Spaced Repetition Memory System part. But, I think @Sascha might have some arguments that I'm still waiting to see/hear on why that shouldn't be necessary.
ok. Something like Logseq, maybe.
I never used this kind of software.
What I've written surely need to be completely reframed, I started considering "notes" as the unit of work of the process.
You certainly have put much thought into your system, I am confident you will get it set up to meet your needs.
To me, though, seeing these notes that are a mixture of folder-like catch all notes, notes with multiple concepts and notes with multiple authors' variants of those concepts - just seems like too much varied information to think about at once. The very opposite of atomic notes.
It makes me appreciate some of the basics of the ZK approach:
1) Source (Lit) notes to capture and store an author's view of a theory or concept
2) Main notes that are concise reformulations of what you have learned in your own words - one concept or theory per note- with your own reactions or comments on that thing
3) Structure notes that can be used to answer a question by linking together related main notes for that topic or purpose (and may include your initial commentary on that question or topic)
So, maybe step back from the folder, tag question and think about the overall parsing of information into Source Notes, atomic Main Notes, and then later Structure Notes. Can use keyword search, keyword index, tagging to get to your main notes. Tagging to solve your need to browse by author or topic.
For me, each concept would get its own page: cognitivism, naturalist reductionist, etc. Having a single page for each concept would bring some clarity. Each concept would have a permanent place to live in your system and could be easily expanded upon, edited, or linked to in the future.
Yes, this is just a single note, with all of the child concepts under it.
No, there are no folders in this image. Anything with a bullet point is a "rem" as @rwrobinson mentions.
I don't know if Obsidian terminology is different, but these are references. I turned "natural properties" into a concept that could be referenced.
Right. I can do that.
Aside from the spaced repetition, it was my impression that Obsidian and Remnote to a certain extent have the same model, since the underlying ideas of tags, references, and folders is still there, along with the graph view. Of course, I don't know much about Obsidian, so I could be wrong.
Right. Cognitivism is the parent concept rem, of which all other rem are children.
Okay I see. If I understand correctly, your main note is the "structure note". In my case, I don't usually have a centralized location of all the connections. It is usually just a bunch of concepts that are interrelated with each other. The example I gave was not very mature though, as I am just starting the book.
Well, I thought your advice was very helpful in getting me to think about how I could use tags and organize my knowledge. The software might be different, but I think the end result is the same:
https://forum.remnote.io/t/what-makes-remnote-an-ideal-tool-for-digital-zettelkasten/1865
I think it would probably work, but I am not sure.
Right. That was my end goal. I am still experimenting to see what works.
And that is where I thought your advice was good. I am going to try going without the main chapter folders, as I think it might be beneficial for making things less rigid.
Well it is certainly nice to find another Remnote user here!
I am not sure what you mean here, probably because I don't use obsidian. What is the difference? With Remnote you can do bidirectional linking through references, and I thought you could do that with Obsidian too.
Your notes look very similar to mine, especially my math textbooks which have a lot of references. You also seem to be maintaining the original chapter structure that the Author uses. I am still unsure of whether I should do this, but I am leaning towards just relying on the concepts to guide the structure and putting in sources for where I find the information, that way I can add in more information from other sources later into this concept structure.
Aaaaah I see. So this is where it becomes more aligned with the zettelkasten approach. I was worried before that would become isolated with connections inside of itself, similar to my remnote math textbooks which have a ton of connections to theorems and such within the textbook, but not many outside of it. If I understand correctly, you are moving it to the ZK folder so that you can connect it to other concepts, kind of building a puzzle piece in the knowledge resources, and then moving that puzzle piece to the ZK folder where it is needed (presumably it has other related information there?)
I am still confused as to what you are doing in the ZK folder exactly. Are you connecting the puzzle piece to a larger whole? Are you saving it there in an organized way so that you can use it in the projects folder?
What do you mean by link source feature? The jargon is getting me here. And when you say that you have Popper's book linked to epistemology, what do you mean by "linked?" do you mean it has a tag with epistemology? Do you mean there is a folder Epistemology that has a portal or a reference to Popper's book?
Right. I noticed that I had a very strong tendency to use a ton of folders, and was trying to move away from that so that I can better use the zettelkasten method. You don't seem to use tags in the image you show, and I am wondering whether you use tags for folders or not. I started from scratch recently because I had things extremely categorized and cut off (I am not giving up my school categories as I need it to be organized for homework/chapters and such).
I am definitely moving away from spaced repetition though, which is not to say that I am giving it up entirely. I got into the community because of the spaced repetition, but now I am having the opposite effect where I want to get more into PKM, but remnote doesn't have that focus. With that said, I believe it has all the tools for PKM, and so I am just sticking with it.
The reason for moving away from spaced repetition is that in practice I find that improving encoding through practice is much more important, along with inquiry based learning, teaching, Kolb's learning cycle, and the like. Also, Zettelkasten is good for enquiry based learning, and research for grad school, which is coming up soon.
Oh yeah for sure. This is in the introduction of the book, and the author was outlining all of the different thinkers and theories, without examining them in any depth. You can think of this as the outline that I plan to fill out, turning some of the bullet points into folders if needed.
This is very uncommon, but the author was kind enough to put an outline of the entire book!
Huh, the "structure notes" in step 3 sound a lot like permanent notes, although from what @andang76 was showing me they seem a bit different. I am only familiar with the Fleeting notes, Literature notes, and permanent notes. The notes I am taking now are the literature notes from the book, which I am writing in my own words as consolidated information. I am attempting to take the crux of what they said in my own words and put it in that skeleton I showed you (sometimes the way they state the crux is so concise I might as well use it instead of my own words). So Essentially I am going straight to "main notes" in step 2, and skipping step 1.
Along the way, I form my own views and opinions, and so start to write up a permanent note explaining my own position while referencing concepts(the bolded bullet points) from the book.
Yep, that was the plan. The whole point of the tags for me is so that I can easily find information when I am creating my permanent notes, or structure notes as you call them.
Yeah for sure. that is what my notes are doing, with making the other rem children to the concepts cognitivism, naturalist reductionism, and such. You can also "zoom in" to each of those bullet points, so that the rem turn into a folder with everything inside of it.
Ok, nice work!
It is clear by now that we have different approaches. Yours is structured, hierarchically organized; mine is a free-floating constellation of independent concepts.
I think this link could be useful for me. I've a knowledge hole regarding how to implement zettelkasten with this kind of software
I would be interested to here why that is, if you don't mind the digression. Does it essentially give the same info as "The logic of scientific discovery?" with the the whole paradigm change stuff?
I know this is off topic, but have you read Hume's Enquiry concerning human understanding? It is one of my favorite books, although I can't say I have read much in the way of epistemology beyond that.
Oh okay. I see. Hmmm. I am not sure whether that is a good thing or a bad thing. Do you use Obsidian?
This makes me wonder whether it is better to have a free floating constellation lol. Like With the skeleton I created it might be too rigid, where it would be better to use tags. I am not sure. Of course it could be changed, but I think that with this structure each rem acts as a folder since it is hierarchically organized. This comes back to the question @rwrobinson was asking me: Why do I think it is better to get rid of folders.
My answer was that getting rid of folders allows you to focus on organizing around tags instead, so that you can have specific content for each project. My ultimate goal is something like this in my permanent notes:
https://zettelkasten.sorenbjornstad.com/#PublicHomepage:PublicHomepage
And he doesn't seem to be utilizing hierarchy, so maybe free floating is better because rather than relying on the hierarchy for connections between concepts, one would be relying on references and tags instead. On the one hand I am worried that without hierarchy I would lose information, and have to use a ton of tags to narrow things down, but on the other hand I don't really see hierarchy in play with any of the zettels that I have seen. I think I just need more examples to help decide.
@StellarThread
Have you read the introduction to the Zettelkasten Method post on this website by @Sascha ?
It speaks to what a structure note is as compared to a regular content note. If you haven't, I'd recommend checking that out.
Are you aware of what a top-level rem is in RemNote?
Actually, I'll just quote from their knowledge base here:
In Obsidian, pretty much, every file is a top level Rem.
I accidently pressed post. I'm going to respond with more...
Yep.
Aaaaaah I see. That clears quite a few things up.
Absolutely. Whatever you like! The reason I do that is just to help see where the concept original came up in the flow of the text. Also, I expect the concepts to move out of there to my ZK section at some point, most likely.
I move it to the ZK section in two main situations:
The ZK is mine. It's imperfect. But, it's my organization and thinking playground. So, when it's important to me and I want to start to engage there more with the ideas/I'm done with the project/processing of the text, I move it into my world.
I think I just answered that. Let me know if that doesn't make sense.
That is a feature. Here is the knowledge base article on it here.
I have a Rem in my ZK section for Epistemology. I use the feature above to link to the Popper book as a source. It's pretty much a special rem reference in my mind. Check out the article linked above.
I only use tags when there is a common structure. I use a tag for my courses and assignments because they have common properities like urls, start and end dates, grading scales, learning objectives, etc. But, yeah, I use tags somewhat sparingly. Primarily for the table features.
I wouldn't recommend moving away from Spaced Repetition. Justin Sung has good stuff. But, I don't agree that Spaced Repetition Memory Systems are only for rote learning. I think, for example, Jeffrey Karpicke's work on retrieval just can't agree with that argument. Sung is right to emphasize elaborative techniques for deeper initial processing, but, retrieval is key thereafter for learning.
Bits and pieces. I think maybe 60%? I need to go through it start to finish.
Can't you just create a portal to view the information when you want? As well as referencing it. Or is it that you only take bits and pieces from sources to narrow down what is important? Sorry for asking so much, I just want to know exactly how you do it, and what its value is when you can access that info wherever you want without having to move it.
I might be answering my own question, but you might be using the fact that it is in the ZK folder to filter for the more important and complete notes.
Oh okay, I thought you were using that term in a special way but it is standard. I use that all the time
.
I see. It seems that Obsidian users use tags a lot more, probably because they don't have the hierarchy that we do since everything is a top level rem for them. I am guessing you fall under the camp that hierarchy is better. Usually, though, when I see a zettel online, it doesn't have hiararchy in the same way. It is more "free floating," likes this:
https://zettelkasten.sorenbjornstad.com/#PublicHomepage:PublicHomepage
I would be really interested to get your opinion on the question of which is better for zettelkasten. You might have done both, so I would like to hear your thoughts there.
You mean this?
https://www.teachingforstudentsuccess.org/episodes/interview-jeff-karpicke-1
Is there any paper in particular that you would recommend from his which specifically touches on the effectiveness of spaced repetition tools? I am convinced that spaced repetition and active recall work, but I haven't seen any studies that look at SRS directly yet. I would also be interested to see it in the context of math, since it is less memorization heavy and more problems based. I eventually found that doing more problems and less flashcards was a better use of my time for learning the content quickly, though I don't know whether this is generally true for people (especially over the course of an entire semester, where SRS benefits from longer time periods).
It was interesting to see that free recall did just as well as prompted recall, and that definitely gives another point to SRS, encouraging me to get back to it. And yes, active recall and spaced repetition are why I got into Remnote in the first place. For context I am a math major. Over time, I used a lot of SRS with my math, but found it to take up a lot of time for not that much yield, and that is despite investing quite a lot of effort:
https://www.remnote.com/a/Introduction-to-Real-Analysis-Fourth-Edition/64713b60fb50020c0d0ae698
I did well in the semester that I took those notes, but that was because I had time. Also it was not long after the semester ended that I basically had to disable the flashcards, probably because I made too many for the procedural parts. In the semester after that I did the same thing but didn't have enough time for it. When I tried Justin's method this last semester without any SRS I made A's in all my classes, and that was with taking minimal notes and mainly focusing on connecting the ideas through inquiry based learning.
My other gripe is this: what is the point of SRS information if you are not going to use it? Eventually you need to disable the flashcards, which is fine, but I still get flashcards from chemistry from years go.
With all that said, I am still interested in incorporating SRS back into my schedule, with the flashcards only focusing on the main theorems and definitions rather than proofs or anything like that(which involves procedural memory. That was another mistake on my part).
Thanks for introducing me to RemNote. I prefer either SuperMemo or Anki, though I haven't been using them with any consistency.
I have at least one subdirectory of my Zettelkasten directory for media. At some point, directory searches could become time-consuming. Splitting a Zettelkasten into separate directories shouldn't be ruled out, if only for performance. @tjex and others incorporate directories in systems that work for them.
So far I have one Zettelkasten directory. If this becomes more of a hindrance than a help, I might see the light and add more, like my experience with structure notes.
GitHub. Erdős #2. CC BY-SA 4.0. Problems worthy of attack / prove their worth by hitting back. -- Piet Hein.
Hello Stellar,
I've been working with Zettel on my DevonThink for 5 years. That being said, I've never found it necessary nor possible to eliminate folders (on DevonThink it's called groups). Folders/Groups are necessary for keeping a Strategic-Operational level organisation of your files and notes. However, you do need to avoid creating a long chain of command. I never make any folder structure longer than three levels. I always put myself on the chair of General Eisenhower, and think about my structure from the military perspective. So I have a group structure represents my armed forces branch, theatre of operations, and armies/air forces/fleets. Now when I want to launch Operation Neptune, I would create a set of smart groups based on tags, built-in metadata and custom metadata to form temporary "battle groups".
I hope my description makes a few sense to you and good luck.