I've tried to organize notes and documents schema-first several times, and it's never worked. For instance, consider a system of notes where each note has tags: I write a note, come up with some tags for it, then I write a second similar note and end up thinking "hang on, I feel like I chose the wrong tags for the first one." So I have to come up with new tags for both.
Either I'm constantly revising how I've tagged older notes, or I have to think so hard about how I'm going to tag each new note before I write it that I never want to write any notes. Before I write anything, I need to think up a set of tags which are specific enough to be useful but general enough to re-use.
Instead, these days, I use Logseq [1], and I'm really liking it. In Logseq, everything is bullet lists. If you have a new note to write, you just append it to your root list. If two notes feel like they belong together, indent them under a parent. If any list starts becoming too bloated, you can move a subtree to a new page.
A list like
Main List
- Entry 1
- Entry 2
- Entry 3
becomes
Main List
- Entry 1
- Category A
- Entry 2
- Entry 3
which becomes
Main List
- Entry 1
- <hyperlink to "Category A">
Category A
- Entry 2
- Entry 3
This way, the schema grows organically over time, from the bottom up. Instead of having to think up a system of classification before I start writing, I just write, and then classify later as information accrues.
Logseq has tags and block embeds and many other features too, but the core nested list model is what has really attracted me to it. I'm sure it's not the only note app that works along these lines (I'm always open to suggestions), but it's open-source and it works quite nicely.
I've tried to organize notes and documents schema-first several times, and it's never worked. For instance, consider a system of notes where each note has tags: I write a note, come up with some tags for it, then I write a second similar note and end up thinking "hang on, I feel like I chose the wrong tags for the first one." So I have to come up with new tags for both.
Either I'm constantly revising how I've tagged older notes, or I have to think so hard about how I'm going to tag each new note before I write it that I never want to write any notes. Before I write anything, I need to think up a set of tags which are specific enough to be useful but general enough to re-use.
Instead, these days, I use Logseq [1], and I'm really liking it. In Logseq, everything is bullet lists. If you have a new note to write, you just append it to your root list. If two notes feel like they belong together, indent them under a parent. If any list starts becoming too bloated, you can move a subtree to a new page.
A list like
becomes which becomes This way, the schema grows organically over time, from the bottom up. Instead of having to think up a system of classification before I start writing, I just write, and then classify later as information accrues.Logseq has tags and block embeds and many other features too, but the core nested list model is what has really attracted me to it. I'm sure it's not the only note app that works along these lines (I'm always open to suggestions), but it's open-source and it works quite nicely.
[1]: https://logseq.com/