Note-taking brainstorming

I was going to reply to the newsletter, but I thought I’d start a discussion instead.
With regards to note-taking, one of the issues I’ve run into regardless of the platform I use is being able to quickly bang out a note on an idea that I have and then integrate that into my structured notes at the same time. If I just try to add it to my structured file, I sometimes lose the essence of the idea by the time I figure out where the idea needs to go within the larger file.
On the other hand, when I used something like Org-Capture to quickly jot down a note, I then have to go to the file it gets written to and move around so it fits within the overall structure of that note. On top of that, there are times when an idea might be applicable to multiple topics, which are stored in separate files which can lead to confusion later on.
Anymore, when I have an idea I need to save quickly, I’ll often just write it down in a physical notebook and then copy that into my org files later on. One thing that has occurred to me is that I might be better off using different methods for different topics.
I’d love to hear how other people approach note-taking and what their solutions and struggles are.


I’d love to hear how other people approach note-taking and what their solutions and struggles are.

I’ll try to outline my approach to this, unsure how helpful it’ll be in the end though.

I use org-roam for all of my notetaking and information-keeping. My process usually bases on:

  • using org-roam-dailies-capture-today (bound to C-c n d n) throughout the day when I want to quickly jot down notes or capture information. TODOs that should appear on my agenda also go into the dailies (all my daily files are added to org-agenda-files) revisiting my dailies file in the evening and then use org-roam-refile.
  • using org-roam-capture for more thought-out things (reading notes, documentation for projects, drafts and so on).
  • relying on rg.el (ripgrep) for full-text search; others than that I rely on backlinks and org-roam-buffer-toggle for discoverability of related notes.

So my org-roam Zettelkasten serves three purposes for me:

  • it’s pretty much a backlog of all the work I do (I use the clocking commands of org-mode on my TODOs, so my dailies also contain information how much time I’ve spend on what task), everything I do as a hobby (there are tons of notes on Badminton, DnD as well as other hobbies) as well.
  • it serves as my personal knowledge-base.
  • it’s my personal source of truth for project management/structuring my work as a software developer.

Regarding Pitfalls:

  • When I started using this approach I created a lot of “category-notes” (e.g. a note “Scheme” where I linked all notes regarding Scheme to and so on). Throughout the years I stopped doing this almost entirely, since the implicit structure created by linking is enough for me most of the time.
  • I used a hybrid-approach at first using org-roam for PKM and a structured folder (using johnny decimal => numbering) with .org files for project management and day planning. I abandoned the separate structured .org folder in favour of using org-roam dailies (being able to link to projects from my dailies has the benefit for me to be able to see what topics I worked on on which days from org-roam-ui, so that’s cool). Would recommend to stick with one system and not doing hybrid.
  • Backlinks solved a lot of categorization issues for me; so instead of figuring out what file to append a specific note to, I’ll just create another note and set links to the related topics in question.
1 Like

Reviewing my captured notes at the end of the day is probably what I should have started doing. I would just immediately go into the captured file and then move it afterwards, but waiting until the evening would probably be a good opportunity to review and think about things a bit more. Part of what I’m struggling with is that there are some topics like ‘art projects’ or silly quips I come up with where a single sentence or paragraph is sufficient, vs an app idea where everything needs to be detailed out and fully developed. I’m not sure if I should work with a system that is flexible enough for both, or use separate systems for managing each type.

Thanks for describing your system, I’ll keep it in mind.

After having tried a few different approaches, I’ve got to a relatively simple solution.

I have an org directory with some fixed, long-running notes files (e.g. a list of movies I want to watch, notes on my home server setup, a list of project ideas, …)

If one of the projects actually materializes, I spin it out into its own file.
Once a project has a git repo or I work together with other people, I put my notes file there or in the wiki.

For mobile usage, I just use orgzly and synchronize with syncthing.

All in all, it works, is relatively low maintenance and I tend to find things relatively easily.


I find the low pressure of the daily journal the only way I’ll actually take notes. I use Logseq (with org file format) on mobile and desktop to keep everything in one environment. I use the the hashtags with the notes so it can auto create pages and back links and all of that happens in my org folder (synced with Syncthing) so it’s usable in emacs ecosystem if/when I want.

But I’ve been very happy with Logseq and really like it’s plugin to bring in articles and highlights from long form reading, a lot of which I do on mobile.
Typical work flow if I find a Linux tool per se:

  • sharing to/pasting a link in Logseq
  • typing in #tools/linux
  • done

Now I can either search (fast) if I know what I’m looking for or go discover to the tools or tools/linux page to see all the baclinks and I can add other tags if I want to but I don’t feel any pressure to refile. There’s very little friction in actually capturing or retrieving information.

Well, my approach is pretty simple:

  • Most captures of random things, like snippets of Web pages, go into ~/org/ (for “Commonplace Book”), which has an Org datetree. I use timestamps, to-do keywords, and tags.
  • Journal/diary-type entries are captured into ~/org/, also having an Org datetree. Usually these aren’t to-do items, but I use tags for pretty much every entry.
  • Information that I think of as “reference” information goes into ~/org/ (e.g. where I put something important, some re-useful information about programming, cooking, etc). But I find myself filing information into this file less often, as it takes more time to file an entry in the right location than I would save compared to just searching for it later (cf. how Web search (e.g. Google) generally beats categorization (e.g. Yahoo)).
  • Things I’m researching or learning about go into ~/org/
  • Captures for, e.g. tasks that I’m likely to do and then not care about in the future, or items that I want to refile, go into ~/org/ in a flat list of top-level headings.
  • I have a few other files, like ~/org/ for info about books, one for movies, etc.

My datetrees go back about 10 years now. File size and performance aren’t issues. (Emacs and Org generally work better with fewer, larger files than many smaller ones.)

For searching I use org-ql-find and org-ql-search (naturally). Before org-ql I used helm-org-rifle, of course.

I use various Org QL views for managing projects, surveying tasks, etc. I bookmark some, use activities and burly for some multi-view “dashboard” configurations, and in some files I also store views in Org links (e.g. ones I don’t use as often).

This all serves as my “personal knowledge base,” or whatever you want to call it. For my various Emacs projects, I usually keep a file in each git repo.

Also, I use org-bookmark-heading to bookmark various entries in my Org files. I still need to upstream that package.

I suppose the principle I try to follow is to let the system do the work for me. Generally I don’t need to think too much about, or even care, which file an entry is in. I just search for it and select the result, as if it were all in some kind of database. I don’t need to spend time filing or linking entries; I’m not building a wiki. Even the tags I assign are often redundant with the information in the entry itself, but I still tag most things I capture. When I do link between entries, it’s between ones that are specifically relevant, not merely topical. And I use Org ID-based links, which are more reliable than heading-based ones.

I guess that’s about it. I expect to use this system for the rest of my life. Why would I want to use anything else? :slight_smile:

1 Like

I have a ~/notes directory in my home directory, and within “notes” there is a directory of random ideas I have had. You could call it a personal journal, though it is not daily or periodical in any way. Every time I have a new idea I just open a new empty text buffer called “*idea*” (M-x b *idea*<RET>) in Emacs. I might set the buffer to Org mode if necessary, the idea isn’t usually there long enough for me to bother with switching to Org mode. I don’t do this every day, so I have no command or key binding defined to do this.

When I am done, if I think the idea fits in to some project, I might save it under that project. Sometimes I might just add it to my TODO list. But often it will just go into my directory of random ideas: ~/notes/ideas. I press C-x C-s, select ~/notes/ideas, save, and I am done. At the end of the day I will commit this file to Git along with any other things I have added to my `~/notes~ directory.

My random ideas directory often serves as inspiration for a blog article. I will sometimes take an idea from there and expand it into a full article.

@ramin_hal9001 You might find this helpful for quickly taking new notes: GitHub - alphapapa/org-notely: Pop to new Org headings for quick notetaking I bind it to M-g N.


Wonderful timing, I’m currently reworking my emacs setup and slowly adjusting my workflow. I’ve been using emacs only since last year summer, and what drew me to it was Org Mode and Org-roam. Then I got hooked on the powers of Org-agenda, and all was well. After a while, I added org-roam-ui, which I don’t really /use/ that much, but I find it a powerful way to discover lone nodes, which most of the time mean they’re poorly developed notes/ideas. I slap an #inbox tag on anything I capture, then filter for any inbox items and try to get through that roughly once a week. My setup is inspired/informed/stolen from @daviwil, Rainer König (I even did the honorable thing and bought access to his Udemy re-recording of the content he’s put for free on Youtube, because it’s good enough to pay for in my opinion). I watched an interview with Rainer and some software engineer that runs a video podcast (that /really/ narrows it down, doesn’t it?) that I found tremendously inspiring.

I did a deep dive into the “building a second brain” (BASB) framework last year, so I apply some of the concepts from there, too. Mostly the routines that surround notes (checklists at the start and end of things, weekly planning/review cycle, etc).

As with everything else, life happened, and I now have hundreds of #inbox entries… but luckily I’m back on the horse and slowly trying to implement one bit at a time instead of adopting an entire system while also trying to learn the ins and outs of it.

I’ve noticed a significant improvement in my ability to retain knowledge, and my commit messages at work have become a lot better (I try to keep notes as I go during the day, when thinking about the stuff I’m working on).

I guess what I’m trying to say is, that while I don’t have a lot of know-how to share in this thread, I’m paying close attention to it as I’m really keen on the subject matter. Thanks for kicking it off, and to everyone for chiming in!