@jethro’s original post mentions “tags” in passing several times, but if I understand correctly these “tags” are just other notes, with no special status. So, for instance, they are not the same thing as org-mode tags. Meanwhile though there is an interesting github issue about tags which I havea bit of a hard time following. It seems to suggests that tags might become something different eventually, a hierarchical structure grafted onto org-roam’s flat-file roots. Am I understanding this right? And also, do people sometimes keep separate stores of notes, maybe say on very distinct topics? One thing I’m imagining is that I will end up changing hte connections in my mind betweeen notes as a project evolves, and eventually I will want to perform some kind of fulltext search across my notes (ag or ripgrep or whatever), but preferably only the notes that have a certain tag. I feel like this conversation has already been taking place but am having trouble finding it. thank you!
I don’t have any complex use cases for tags. Ahrens recommends in How to Take Smart Notes, to have different kinds of notes:
- bibliographic notes
- Project notes
- Fleeting notes
I intend to have tags for “refs”, and 1 tag per project. This is solely for me to aid filtering by biblio notes, or by project on the interactive commands. So yes perhaps you put it right: I’ll be using it to introduce categorization on an otherwise flat structure.
Lot’s of good content on the zettelkasten.de site. Here’s a couple of posts that deal with tags:
https://zettelkasten.de/posts/object-tags-vs-topic-tags/
https://zettelkasten.de/posts/luhmann-folgezettel-truth/
cool, helpful, thx. but do tags in org-roam have a special ontological status as of now, or are they merely notes with the same properties as other notes? And also, are there plans, e.g. in that tags issue, to change the ontological status? I had a little bit of trouble following the conversation…
thanks, very interesting posts.
The tags you currently see in O-R have no special status; they’re just lists of O-R links.
The new PR does formalize tags though.
I’ve written some docs here:
Let me know if I’m missing out some use-case.
I think the docs look good, as they make clear that the primary purpose is to categorize (or type) notes, and NOT to assign meaning to the content of those notes.
I could imagine this could be a source of confusion, but I think the docs point people in the right direction.
@titaniumbones - one reason the tagging conversation is confusing is that Roam implements tags differently to most systems. Tags are typically a separate property from note title, but Roam takes any tag, creates a page with that title, and backlinks all notes that have the #tag in them to that tag’s page. When they let you narrow a group of notes to those with 2 tags, it is just a search for notes that link to both of the 2 tags’ pages.
I just saw the discussion/update here and have a question. The feature says that tags can be “obtained” in 2 ways. But how are these “obtained” tags displayed? Or, how does the user actually interact with these tags that can be indicated in 2 different ways? At first I thought the idea was to scan a note for the subfolder and then add the folder name to the note header as a #+ROAM_TAG, but now I don’t think that’s the case.
Just wanted to mention that as a beginner I found this overloading of the term ‘tag’ confusing. It means something very specific in org-mode, so I assumed it meant the same thing in org-roam. It would be useful to clarify that in the docs, or maybe use a different term for these links.
Yes it seems to be a recurring theme of confusion. Perhaps roam_tags should be renamed roam_labels or something else?
I don’t see Roam behaviour as confusing, in fact, it’s standard wiki practice: before having real tags we had the links that started with a reserved keyword (usually “Category”).
So -as an example- if you were talking about Buenos Aires city and added a link to CategoryArgentina, then a page named CategoryArgentina was created that listed all the nodes (or notes in our case) that had that link (the equivalent of being marked as belonging to that particular category). And, of course, there was a page (node) called CategoryCategory that listed all categories. 
Actually, this kind of configurable backlink search is something I’d still like to see implemented in O-R.
As a simple example, imagine you are writing about system thinking and global change and had tags about them (SystemThinking and GlobalChange). You may like to end this note (or node, in hypertext parlance) with a list of all the notes marked with both tags.
In traditional wiki engines that would imply some kind of object or markup that will create that list in the page (what we have in O-R in the default side pane of backlinks, but configurable on each instance). In MoinMoin, for example, they are called macros (see <<FullSearch()>> in http://moinmo.in/HelpOnMacros) and in this example it would imply to search something like <<SystemThinking+GlobalChange>>.
Of course, this macro works because wikis works on 2 modes: edit and reading, so the macro is expanded in real time when in reading mode.
However, even if the mechanism may need to be another, the function/service is really good: you can create lists comprised of related notes as arbitrary searches and include them where they are deemed useful…