Did you try https://logseq.com/? I've had not reason to pick freemium Obsidian over open source Logseq.
That said Obsidian and Logseq are interoperable since they both run on a local folder of plain text files. Meaning you can switch over to Logseq for your outlining needs and use Obsidian for everything else.
(slightly biased since I helped on Obsidian 1.0, but I am a lover of all plain text tools)
Often when I write down my thoughts this way, it is more like following associative threads. I focus on a particular thought and relate something to it, which now becomes my new focus. Then I defocus and focus on something completely different. Kind-of like these threads here on HN. I wouldn’t call this outlining, it is more like spawning local contexts, nested textual environments to think in.
This is something that I miss the most when working with linear text structures as in Obsidian. I know you can indent and fold indentations in/out, but for me it doesn’t feel natural the way it does in notebook apps that organize text in block-trees. I also cannot reorganize or reference those indentations easily. I feel liberated (in thinking) with those bullets rather than constrained, but of course it is a matter of personal preference and habits.
I’ve used dozens of outlining and mind mapping tools from FreeMind to Logseq to OmniOutliner and more… But for me Obsidian, still wins because of the plugins. Check out the selection of outlining, link management (in particular link graphing) and crucially the refactoring plugins.
At the basic level the outlining plugins give you shortcuts keys to rapidly realign, reindent, fold and navigate a large tree of indented text, and when you combine that with the ability to take an entire level of that tree and just slice it out into its own document, leaning a link to the new document in its place (which other plugins can use to display an inline preview of that document) … it’s just amazing.
- For fans of "outline workflows" Outliner is excellent. A whole bunch of outline/indented text movement and manipulation commands: https://github.com/vslinko/obsidian-outliner
- For easily refactoring notes that are getting too large you want to have Note Refactor. It gives you tools to easily take blocks of text and quickly cut them out into new notes. Its not magic out of the box, but its a powerful tool you can use when building workflows with other plugins. https://github.com/lynchjames/note-refactor-obsidian
- Local images is another good one, working with online content can get messy when you copy notes and then want to be able to work any where you have Obsidian synched. I've got it on my Laptop, two desktops, phone and tablet... I want to carry as much of my related content with me so having an easy way to convert remote images to local copies is a big productivity boost when making notes about content from the internet. https://github.com/aleksey-rezvov/obsidian-local-images
- For analysing the content for some useful stats there's: https://github.com/SkepticMystic/graph-analysis but this is for a relatively specific sort of analysis.
- More general and flexible analysis and graph visualisations are available from the combination of https://github.com/zsviczian/excalibrain , https://github.com/blacksmithgu/obsidian-dataview and https://github.com/zsviczian/obsidian-excalidraw-plugin ... in short query your notes and note metadata like its a database, build reports and data visualisations, and then excalibrain is a whole thing built on top of that power.
- Dynamic embeds of outside content are available from https://github.com/dhamaniasad/obsidian-rich-links and https://github.com/Seraphli/obsidian-link-embed depending on the style and use you like. While there is a built in functionality to preview the links to other notes when you hover over them https://help.obsidian.md/Plugins/Page+preview which has a demo here https://youtu.be/dmnVml_jbsQ?t=222
- And a real force multiplier is adding https://github.com/Taitava/obsidian-shellcommands to your setup. It lets you run scripts and prompt for information and really invest time in procedural automation without having to build your own javascript plugins. So you can setup your system so that when you use the refactor to cut out a new note, the automations will trigger, ask you to give the note a new heading, tags, and you have a little script that checks last modified time of the folder tree of text files, and looks at the folder of the last modified time and asks you in that popup if you want to move the new note to the folder the note you cut it from is located in. Or anything else you can imagine using outside automation and scripting tools on your plain text markdown files.
These are just a start and if you haven't already browsed the plugins at https://obsidian.md/plugins I wholeheartedly recommend it, people are adding new cool things pretty often and other plugins add new functionality that makes them worth checking out if they were previously not something that you found interesting. I do a read through of the plugin list probably at least once every month or two just to see what's new, and more often if I'm experimenting with changes to my workflow.
So I think while outliner might be misplaced for emacs it is not for logseq.
However, I think Org-mode’s headings (*, **, …) have great nesting and folding capabilities that are more similar to the structure Logseq or Roam provide, but I don’t like having to create a heading hierarchy for my thoughts, so its not really the same.
My critique was more that the term “Outliner” does not really describe how I work with tools like Logseq and Roam, since I don’t really outline my notes/thoughts but use nested bullets more like a focus hierarchy that helps me to “anchor” them in the context of a previous note/thought (I hope this makes sense).*
I wanted to use logseq (I felt good about "Obsidian, but open source") but when I tried to find some text in a page I was writing it didn't work. I'm a total logseq noob but as far as I could tell I needed to install an extension/plugin to search through the page, which was weird. Plus, the plugin didn't work for me (I typed in the thing to search for, clicked "go find it" and nothing happened, I think - it's been a while and I didn't use it much).
I kinda boggled my mind that logseq wouldn't have a 'Find' feature for finding text in the page I'm editing.
Please tell me that I missed something obvious so that I can feel dumb for missing that obvious thing but happy that I can take another look at logseq :)
The UX is extremely lacking, but it's open source, gaining steam, and they recently closed $4mil in funding so I expect it to massively improve over the next two years. Notably, page search should function without needing a dialog+overlay, and should support highlighting/navigating every match.
I'm excited to hear that there's support for expanding it.
And I'll definitely go back and take another look at it! :)
I think I tried Control+F, and maybe looking through menu items (it's probably there & I just missed it).
First note taking tools where I actually *read my notes*
Enter Logseq, and after a 20 minute learning curve, ideas just fly off of my fingertips. I reach for it daily. Can't recommend Logseq enough.
When you open Logseq, you start with bullet points. Is that the only thing that pushes you to create more? In Obsidian, you can just start bullet points on your own.
Nits aside, I use both and sometimes folder structure came in quite handy (like having separate notes for course modules and having 15 topics). I wouldn't remember even the names of these topics to come by when I need to.
But I agree also that the magic in Logseq happens more often than in Obsi (rediscovery). I think what contributes to it is the atomic nature of blocks as opposed to pages and daily scroll of all topics encountered recently.
While waiting for Logseq to come to Android, I've been using Zettel Notes by Dev Rohit. It's been great so far!
what? why... also emacs is really good for reproducible terminal logging of experiments/commands/output.
org mode has been too time-consuming in the past to adapt into my workflow. Logseq is a fantastic compromise.