Forum Posts Feed
The Benefit of Nested Tags
< Next Topic | Back to topic list | Previous Topic >
Posted by Amontillado
Dec 17, 2020 at 07:05 PM
I use tag hierarchies all the time in Devonthink, but I discover I have a question.
If I use the same tag name in two different parent tags, how does DT choose which one to use?
The answer isn’t jumping out of the documentation. It appears to use the most recently created instance of a tag.
For instance, consider two tags called “x” and “y”. Make a tag “a” under each. When you add a tag “a” to a document by typing the tag name in the tag bar at the bottom of the document, which “a” gets added?
Inquiring minds want to know. I think I’ll post on the DT forum.
Posted by Amontillado
Dec 17, 2020 at 07:34 PM
Ok, answered my own question.
My guess was correct. The DT manual says that adding a tag goes to the most recently defined instance of that tag name, if the name appears more than once in the tag inventory.
DT kind of mildly suggests tags should be uniquely named, but for any who like to use the same tag name in multiple hierarchies and you want to use an older version of the tag, drag the file from the list to the tag you want in the sidebar.
That way I can have tags Opus 1, Opus 2, etc., each with subtags Chapter 1, Chapter 2, etc.
I like that. Notes can appear in whatever Opus/Chapter tags that make sense.
Then, when I come to my senses and realize I’m presenting things in the wrong order, I can just make an Opus 3 tag and re-arrange without throwing away what I came up with for Opus 1 or 2.
Works great, but DT still has a point about avoiding duplicate names.
A document might have tags “Opus 1 Opus 2 Chapter 1 Chapter 1”. If you want to remove the document from Opus 2’s Chapter 1, you need to go into the tag for Chapter 1 inside of Opus 2.
Delete the document there - deleting in a tag just removes the tag, it doesn’t delete the file.
You might think deleting the parent Opus 2 tag would do the trick, but if you delete a parent tag while leaving a child tag in place, the parent will reappear. I assume that’s from the way parents are automatically added when a child tag is applied.
In my use of tagging to create alternate grouping, I have been using descriptive tag names. I think if I start using generic names like Chapter 1, I’ll make the tag name 1-Chapter 1, for Opus 1, Chapter 1.
Or something like that.
I would apologize for obsessing over nittery, but it’s what I do. Seems perfectly natural to me.
Amontillado wrote:
>The answer isn’t jumping out of the documentation. It appears to use the
>most recently created instance of a tag.
>
>For instance, consider two tags called “x” and “y”. Make a tag “a” under
>each. When you add a tag “a” to a document by typing the tag name in the
>tag bar at the bottom of the document, which “a” gets added?
>
>Inquiring minds want to know. I think I’ll post on the DT forum.
Posted by Lb
Dec 18, 2020 at 02:26 AM
Pierre Paul Landry wrote:
LB wrote:
>> It seems Nested Tags and Hierarchical Tags are the same to some of us
>
>Indeed, the two can be used interchangeably.
>
>My previous post sounded more “sales-pitch” than I originally intended.
>Sorry. The point I wanted to get through is that not all tagging systems
>are equal, and in particular, nested tag ones.
>Pierre
>
Not sales-pitchy at all. In fact I’m a big fan of InfoQube and out of all the IM programs I use it’s the main one and probably only one I think that does Tags right along with most everything else.
In my thinking, Nested Tags and Hierarchical Tags are the same. I wasn’t sure if that was correct on my part though since a search for “Nested Tags” has a lot of results for HTML code, which I’m not sure is the same.
Posted by mkasu
Dec 18, 2020 at 04:40 AM
I’ve played around with Obsidian quite a bit, but ultimately ended up switching away because all this loose structure did not work for me and I started to lose files in the graphs because I forget links or tags.
I since started to use Bear and nested tags quite a bit.
For the hierarchy, I mainly enjoy that I can fall back to parent tags. I am academically supervising students, so I might have a note tagged with “#people/students/John”. I can easily filter for “John” to get his specific notes, but also for “students” to get all student notes, or for “people” to include notes related to coworkers or professors. I have the same for #projects/ProjectA and various other things.
I also give each notes multiple such tags, so a folder structure in Obsidian wouldn’t work (It gets ambiguous; would I put a note on John’s ProjectA in the project folder or the students’ folder? What if one student has multiple projects, what if one project is done by multiple students…)
I suppose, I could use multiple normal tags and filter for combinations like “#students AND “#john” or something, but without the hierarchy lots of structure would get lost very quickly. Also I might sometimes forget to tag one or the other. The hierarchy forces me to include parent tags automatically.
My main database for archival is DevonThink, though, so all my PDFs etc rest there and I use external links in Bear to link to them. I am currently trying to mitigate as many notes as possible from Bear to DevonThink in order to make use of all the new backlinking features, but I struggle to implement a tag workflow in DT comes close enough to the way I use Bear tagging.
Posted by Christoph
Dec 18, 2020 at 10:08 AM
For me, the less rigid structure is what makes Obsidian so attractive. It is not strictly neccessary that everything is connected via links and tags, you can find content easily even without that using normal text search and the “unlinked mentions” section. I also make use of nested folder to structure my Obsidian vault.
mkasu wrote:
>I’ve played around with Obsidian quite a bit, but ultimately ended up
>switching away because all this loose structure did not work for me and
>I started to lose files in the graphs because I forget links or tags.