On the sustainability of folderless notes structures

Splitting out an interesting conversation on folders and notes into its own thread.

I’m quoting @alltagsverstand’s initiating reply and then moving the other posts in below:


Text search would work
Regex would work
Tags would work
Even links would work using search/regex.
Folders are a choice, naming structures are a choice, and neither is frictionless.


I don’t say that folders don’t make any sense at all - I am just saying that you are not that dependent any more on folder structures than you were with a phyisical environment or with just using file managers.

I use folders in my vault as well, separating for example between work-related notes, private organization stuff, zettelkasten notes etc. But, for example, within my zettelkasten, I don’t use any subfolders as this would rather restrict my productiviy instead of triggering it. But all of these workflows are very subjective of course - that’s why I started by saying that it all depends on what your actual purpose is.

And, as @Dor said, even if obsidian wouldn’t be available anymore somewhere in the future, with tags, internal links and so on you can just open your notes with a bunch of other markdown editors - and you can use many other tools as well in order to bring a new structure into your files.

1 Like

Which is actually great as exactly because of this you don’t need any apps - you can just type a very simple line into your terminal, which would for example allow you to search for all tags named for example #tag even in a million of files and moving them automatically to a separated folder - all of that within some seconds… Try to do the same by browsing through your local folders! This is actually the huge advantage of plain text files - you are not dependant on any apps at all…

Actually I don’t really get your point… What do markdown editors have to do with being online or offline? :thinking:

No, any program that can read text. Markdown is irrelevant for this.
I can do this offline and on mobiles.

No. They’re just a text string. But a curated or semi-curated text string which adds extra power to a whole text search.

That’s the point. It’s a choice. Some people like folders. Some people don’t. Some people are very good at advanced regex type searches, others aren’t. Some files fit comfortably into one box, others don’t.

I have got to ask. What was the domain? Why were these files so important?

Related… I think three things are missing from this conversation: context, scale, and cost.

If you’re doing work that may have legal or historic implications (and I suppose anything could have either of those eventually—the world is a weird place), you should be taking scalable, sustainable stewardship into account.

Scalable means “Does this organizing approach still work if I have orders of magnitude more items than I currently have?”

Sustainable means “Will this still work in 20 years, with whatever technology we might have?”

Stewardship means “Will this work for others, if I have to get someone else to work with this?”

This is where context comes in. Many of my notes are just for me, so I’m not so worried about these questions.

And finally, this is where cost hits. Any organizing schema has two kinds of costs: input (storage) and output (retrieval). The more rigorous and complex a system is, the more likely it will be costly to file something away. However, it’s also likely easier—less costly—to retrieve it. For your average “note-taker,” balancing these two costs is going to be paramount.


I have the feeling we are talking about completely different things… Of course obsidian (or any kind of text editor) is not meant as a system that manages all your files - in the end it is just a note-taking program… And notes, in the end, just represent my thoughts on different topics - with the final aim (if you use the zettelkasten method) to link different thoughts together in a - sometimes - unexpected way.

Actually, this is what happens when someone who thinks in folders encounters information that isn’t in folders. It doesn’t have to be like that.

I have some very large archives and I use others which are many many times larger than that. Some are in a highly complex system of folders and some are essentially random.

With undigitised paper documents there is no choice. Each document will be in a box on a shelf in a bookcase, and there will be only one way to find it. Even if there is a highly structured logic behind the original structure, it needn’t make much sense now and will often not be intuitive.

With digitised documents, I never utilise any folder system because search will be much faster. The folder system will still exist, because that will be the guide to the placement of the paper document, but I won’t use it.

With pure text files that is even more the case. And tags, depending on how they have been constructed, add to the ease of search.

If I discover an old device or memory stick with potentially useful files, I never attempt to merge any folder systems. I might remove duplicates, but that would be it. With this system I can find any file I want, I can check whether it exists or not in seconds (which would tend to be very slow in a pure folder system).

I have shifted computers, devices and operating systems quite often. I’ve found it is much faster to use new search programs and start from scratch rather than try to move an existing index.
With these the biggest time cost is the analysis of more complex file formats (docx, pdf etc etc) to create an index; plaintext is fast to process.

As I said, I am not saying this is how people should do it. Some people will find it much better to work with folders. But it’s not the only way.

I don’t think Google keeps the internet in a series of folders.

[quote=“StoltHD, post:10, topic:7619”]
if you need to reinstall your OS or move files and folders to another computer without the needed software, you shall be able to find what you are looking for just by browsing [/quote]
Well this might be true. If I should ever find myself with the files and a computer that only has a file explorer that can’t search.

But I think my planning will be done on the basis of having computers that do have a search ability.


It is also possible to have both, with “Link Shell Extension” …

  • a loose note space, without structure,
  • and a structured folders system with “physical links” and/or “symbolic links” etc, to the respective files.

Thus the files in disorder are moreover accessible through a system of folders.