all of my knowledge is encoded in my open browser tabs

I’m rarely more than a few steps away from a notebook and pencil. In my experience, the act of writing the notes that matters far more than reading them - most of my notes are never read, but the act of writing focuses my thinking in a way that often helps me remember.

For a very long time now (like 20 years) I’ve adhered to a rather strict regime in which I write things down. Not as in “jot them down and never look at them again”. I mean fairly careful write-ups – some of them are more like research reports or literature summaries, others are more like lab notebooks, depending on what I’m doing, but they’re pretty substantial.

E.g. If I read a paper, I will do a standard(-ish) write-up with a summary, a discussion of the conclusions, anything that I’ve reproduced or failed to reproduce, directions for further research and the like. If I read a book that I find particularly interesting and annotate a lot, I will write a report of sorts, including a detailed summary, a discussion of the author’s main points, references to other works and so on. It sounds like a drag but really it’s not: it forces me to re-articulate everything I learn, put it in context, explain it in my own words and so on. If I take notes as I’m writing a book, it takes me about an hour or two to write the “book report”, so to speak, and it’s immensely helpful. I don’t do that with every book and every paper I read, just the ones that I think I’ll care about in the future.

I reference these things pretty often over the years, and some are under active editing pretty much all the time.

The “setup” I’ve used in the last two years is pretty uninteresting. I think the ones I tried may be more relevant because of why they didn’t work. What I use now is pretty boring, it’s Obsidian, with three caveats:

• Except for Zotero integration and paragraphs anchors (and I have migration plans and the convert-to-an-abstract-representation scripts for both) I don’t use any Obsidian-specific Markdown features, it’s all vanilla. This is mainly because I’m concerned about vendor lockdown: some of the material in my Obsidian vault is 20 years old at this point, and has outlived the application it was originally written in (Netscape Composer) only with significant time investment in converting them and I’d rather not do too much of that.
• I use a bunch of CSS hacks because the default Obsidian interface takes up a lot of space and is in that weird spot where it is, at once, retina-scorching bright and with surprisingly little contrast, which makes it hard for me to stay focused. I wish I could do something about those awful icons but, you know, one impossible thing at a time.
• I don’t use Obisidian’s document-title-as-file-name convention. They’re all regular files and I stick to my own naming conventions for them. This is, again, mainly due to vendor lockdown concerns, but also because some of the stuff in my vault was written on Windows 98 and moved around through Windows 2000, FreeBSD, Linux, OS X and macOS systems, which has taught me to carefully avoid anything that’s not alphanumeric in file names :-D. But it also has a nice aspect – the whole thing can also be easily browsed and edited with just a file manager and a Markdown editor, even over SSH, which I sometimes did pretty often.

I synchronize it on a handful of machines with Syncthing. I found that anything which requires more manual action than that (it used to be FTP, then git) or relies on permanent network access (SMB) just doesn’t work in the long run. Similarly, anything that has too many moving parts (MediaWiki) is just not for me.

The whole thing is organised pretty much like a website. It actually started off as an offline website long ago (hence Netscape Composer). I write things so they’re easily grepable, and I do rely on Obisidian’s search feature sometimes, but I mostly get around by doing the clicky-clicky.

Over the years, it has gone through several technologies. Some stuff did get thrown away in-between transitions though – mostly things that I thought were just of historical interest and I’d never have to reference again. I have it backed up (save for a few years’ worth of data between 2003-ish and 2011-ish, much of which got lost in a backup accident). It’s been through:

• Netscape Composer
• Didiwiki, then MediaWiki (bad mistake, took me a while to figure it out)
• LaTeX (still didn’t learn my bloody lesson)
• Paper notebooks
• Plain text files with manual filing
• werc, then a clone of werc that I wrote (finally learned my lesson)
• Emacs with markdown-mode
• My own hypertext system with WYSIWYG editing
• And finally Obsidian

Why Obsidian: it’s the most straightforward tool that a) edits a language that’s easily converted to and easily enough converted from if I ever want to migrate to something else, b) has both easily-accessible hyperlinks and decent WYSIWYG-ish editing and c) has an editor that can actually handle what I write.

b) is a big deal for me. I’m all for separating form and content if I’m typesetting a book or writing documentation for someone else. However, editing markup language interactively while I’m actively working on something else is a drag and I hate it. That’s the main reason why I abandoned werc. Doing all my writing in Markdown with acme felt like peak l33t m4d h4x0r but it was very unpleasant for anything longer than 700 words and pretty much everything I write ends up longer than 700 words.

c) was pretty depressing honestly. I have an operating systems “notebook” with notes about operating system design and implementation that dates back to the Netscape Communicator era. It’s, uh, pretty big (338411 words by wc‘s latest count) and that’s just the “main” file, which has tables, images and the like, plus links to programs and whatever.

Most note-taking applications out there will crash on it. If I load it into Google Docs on a machine with 32G of RAM the typing lag is about 10 seconds. It will get a Macbook M1’s fan to spin. I’ve found Obsidian was the only recent application that could handle it.

(Edit: why depressing: just for shits and giggles, years ago, I tried to get Word 97 to deal with that on an old Pentium 233 beige box. Barring import artefacts – Word 97 doesn’t read straight Markdown files so, uh, it wasn’t the easiest thing – it worked like a charm. The file took forever to load but once it was loaded, the whole thing was super snappy. On a single-core 233 MHz machine with 128 MB of RAM. But hey, maybe 2024 will be the year web technologies will finally close the gap with native, then the year of Linux on the desktop is sure to follow!)

Why not the other ones:

• MediaWiki: annoying to set up when I change computers, migrating things from it is a pain
• LaTeX: no WYSIWYG editing, sue me
• Paper notebooks: I used that a lot between 2009-ish and 2015-ish. I now have a large box’ worth of papers and notebooks under my bed. My backup strategy for it is well I hope they never get wet. Also very hard to use for code which is… kind of important to me, what with being a programmer and all. Editing anything that I wrote was difficult. That’s why they were never “real” notebooks – I used a bunch of mechanical notebooks/folders, so I can easily add new sheets as I needed them.
• Plain text files with manual filing: the filing was a pain and it caused me to cluster things in single documents rather than spread them around. Also, I actually do need to archive images, videos and the like sometimes.
• werc: see LaTeX above :-D
• Emacs with markdown-mode: limited support for embedding images and the like. I’d badly hacked markdown-mode to handle images more gracefully and support anchors (so I can easily cross-reference paragraphs, not just sections, within the same document or between files) but it was quite the effort to maintain it.
• My own hypertext system: that was a lot of effort to maintain.

The disappointing parts:

First, it’s just way too much technology and I hate it. If I didn’t have to reference so much electronic material (read so many PDFs, write so much about code, annotate so many binary dumps) and maybe if I had a bigger house, I would probably just use notebooks. Modern editing systems are very restrictive (see below), the ones that try not to be restrictive are just whacky (mind maps and canvases, seriously?) and even low-maintenance solutions like Obsidian accrue a lot of wasted time on account of dealing with general computer nuttery, like software updates and whatever.

Second: since I’m bound to electronic notes for my sins, what I really want is just Netscape Composer with UTF-8 support (so… Seamonkey Composer?), better math handling, and better structured editing (so I can easily navigate between sections by doing the clicky-clicky rather than doing Emacs-style search-forward or going page-up/page-down).

I find the lack of formatting options in Obsidian very annoying and disappointing. My real-life notebooks had well-integrated drawings, colours, various writing conventions for various types of notes. I can’t do that with Obsidian (or any modern tool, for that matter), mainly because they take a convention from the world of typesetting (separating form and content) and try to apply it to a field where it really doesn’t apply. So I get beautiful, but largely dysfunctional documents, which conform very well to the sophomoric “always separate content and form” mantra but are about half as useful as they could be.

My old Netscape Composer “website’ had all that and it was by far the most useful tool I had for collating information. However, it just wasn’t meant for editing large documents, so it lacked a lot of things, like navigating sections, creating new files on the fly and stuff. I could replicate a surprising amount of that with my own hypertext system but it was just a lot of maintenance burden for just too little reward, and at a bad time in my professional life, when I really couldn’t afford spending time writing non-work software.

Currently I use Fossil’s wiki. I host a copy running on a VPS and I run it locally also. This is definitely not it’s primary use-case, since it’s a full-blown VCS. My biggest problem with it is on Mobile, the editor is not the most friendly and it requires a decent connection(since it’s just a website, with essentially no JS).

I like org-mode, I want it to succeed for me, but Emacs is just annoying enough in subtle ways that it’s too annoying to be usable. It doesn’t integrate well with any platform, it’s to dedicated to being it’s own platform. Which might be fine, if you could use it everywhere, but Emacs and Mobile are not friends and never will be. I usually try again every few years, but it just never works past a month or so, because of Emacs.

I have to use paper and pen. Computers are too distracting. As soon as you sit at one, you forget what you were doing and just do whatever is on screen instead.

The main feature for me on the hardware/software side is a smooth sync experience. I am always near a device, but the one I use for capture is rarely the one where I need the content. Beyond that I don’t require much and value simplicity. For now I have settled on Joplin and while there tons of ways to sync with that I paid for their cloud service because it is easy, it helps me sponsor the project and they have e2e encryption, so I don’t mind putting my personal info in.

Much more important, however, is the process and the discipline to follow it. You can dump a lot, but then you have lots of random snippets lying around scattered and what are you going to do with that? There are good approaches to transform that rambling into a useful set of notes. For instance the approach Sönke Ahrens describes in How to Take Smart Notes

But, I always struggle with the balance of how much effort I should put in and what benefit it brings. On one hand having easy accessible and good notes saves me a lot of time. On the other hand it takes a lot of time to make it work. Does it save more time in the end? I am still not sure. So sometimes I just quit and then a month later I really want some info that I know I recorded, but threw away when I quit and on the spot I start all over again.

Somewhere in the future I see a personal LLM help me with the tedious part of this process somehow, but we’re not there yet. i think it is smart to embrace the chaos a bit and just accept that it is never going to be “perfect”.

As an aside, I have an Onyx Boox Note Air 2 e-reader/Android tablet and the handwriting recognition is amazing. I have horrible handwriting too, but to my surprise it works almost flawless. The major downside is that only works that well if you use the official apps, which are tied to their own sync service. Also it is just a fraction too slow to be really comfortable, but otherwise, I am really impressed.

So lately Anki has actually been a bigger and bigger part of this for me.

The first point of contact is usually a syncable markdown thing like Obsidian (if I’m on my laptop and especially if the information is verbose) or my iPad (I preferred writing on paper but never had it available when I needed it; so on an iPad I get a nice balance of handwriting plus better organization and ofc. I can use it directly for books or even Obsidian).

However, I think we’ve all experienced our memex suffering a kind of bitrot. I’ve gotten much better at the art of putting information into Anki, and now I move stuff from my markdown/written notes into Anki where applicable.

I used to spend a fair amount of time organizing my notes and thinking about how to optimize it for reference and matching my mental model, etc. etc. But then when I started just memorizing more of it instead all of a sudden that whole class of problems went away.

From what I recall (ironic?) the lifetime cost of memorizing something via spaced repetition is about 30 seconds, so a very large number of things, but not everything, become economically worthwhile to memorize. Tomes of potentially information I usually leave as data. However, I’ve sometimes found it useful to use Anki to remember where to find information, and make sure I “know what I don’t know”.

Another thing I’m interested in trying out soon is getting a decent self-hosted language model like Khoj running. I’m hoping I can just more literally dump a stream of consciousness into it and it will be more effective at drudging things up and avoiding that information bitrot I mentioned. It might also do a decent job at Ankify-ing things on-demand.

Either way the goal has been: minimize the friction of writing something down in the first place, then minimize the pressure on my knowledge base itself to usefully store information for long periods of time, either via memorization or (maybe) a LLM.

I will say though that while I’m grateful Anki exists, I do think it could be a lot better software-wise, so maybe one of these days I’ll work on something similar. The idea of a memex that’s the intersection of all these things is rather interesting.

(Side note, for those who don’t know marksman is a language server for markdown so you can basically have Obsidian in your usual code editor if you prefer).

VIM in a tmux session over SSH, with automated back up, including link targets. Sometimes paper for a change (or a drawing tablet, or a phone with a stylus), but it ultimately ends up in a text file, which is by far more convenient to work with. I rarely feel a need to draw.

Analog:

• I try to keep an A5 ring bound notebook in my backpack, blank pages (possibly dotted). I use it mostly during meetings where I don’t have a digital device, and for jotting down things that aren’t just pure text (graphical, relational, mindmap-ish, etc)
• I almost never digitize the contents of my notebooks, and I rarely read it later. I guess it’s mostly an aid for thinking while jotting things down, and for adding more associations to help my brain internalize it.

Digital:

• Notes.app: One of Apple’s best apps. Synced, support for all sorts of contents, pen input on iPad. I often use it the same way I use my analog notepad, although the sketching experience isn’t as nice even with a Paperlike screen cover.
• Obsidian: Started using this for more structured use, for things I expect to need later on and especially for work things. I like the Markdown support, and the support for Mermaid JS is very useful when describing state machines and sequences. The fact that storage is a bunch of Markdown files in a folder gives peace of mind, and flexibility to edit in creative ways when needed.
• Things: On-and-off use as a todo list, although I find it much easier to add things than manage and finish them.
• TextMate: “New file”, enter text, keep around as untitled file until no longer needed or save as a “real” file if it’s something that’ll be needed later.
• Web browsers: like someone else said, my open tabs are a part of my way of keeping track of things to look into. For Firefox I use and rely on TreeStyleTabs, in Safari I use the tab group feature.

Sync is an important feature, as is multi-platform support (in my case I consider Apple’s platforms to be multi enough: Mac OS, iPad OS, iOS).

I would never consider putting things like this in a web service, and try hard to keep to sync options that reply on platform features instead of relying on random cloud services that may or may not stay around.

I write things on paper. Paper can’t be hacked or run out of battery.

I take notes on a reMarkable device that I bought for grad school. It’s basically like dumping to paper, except it has undo and a bunch of other features, and what I write ends up synced to my computer. It’s honestly really nice, though I’ll admit the device itself is overpriced and so I’m not sure I’d recommend it strongly for that reason.

Have you looked at Cosmo? I’ve been wanting to try it, but the non-ANSI keyboard put me off. They seem to be real (there’s a healthy supply on eBay) and sell devices with tons of layouts.

https://store.planetcom.co.uk/products/cosmo-communicator

Specifically for stream of thinking (e.g. first draft of some explanation) I have used voice recording (what device cannot do it nowadays) and fed it into Whisper. I used English, which is not my native language, Whisper worked fine despite my accent.

(For small notes that «clearly should be taken» I use physical-keyboard-equipped devices, handheld when not at my laptop, see https://lobste.rs/s/xaj9bz/what_s_your_braindump_second_brain#c_gms58s for details; I probably should take more notes, but I lack a workflow.)

Capture:

• For quick reminders — a new restaurant I just went by, a thing I just remembered I need to do, an item I need to remember to buy — OmniFocus inbox, accessible from my iPhone lock screen. Always with me, always able to capture something brief. On iPad or desktop later, I can organize these into actionable thoughts, or check them off if I otherwise captured them somehow (e.g. somehow already remembered to do the thing when I got back home).

• For notes — Apple Notes, typically on iPhone or iPad. I have a system of 10-12 folders, and by default, things go into the generic Notes folder which Apple starts you out with. Later, at my desktop, I can see that there are unfiled Notes which can be further organized or captured into other apps if needed.

Brainstorming / thinking:

• I’ve recently become enamored of Muse, primarily on iPad. It’s designed around sitting back, typically not at your desk, and freeform thinking, collecting, lightly arranging, doodling, and a lot more. Later, at my desktop or with the iPad flat, I can organize, further notate, or review items.

There’s a number of other more purpose-driven tools that I use, but these are the main ways that I make sure that I don’t lose the thread on anything that pops into my head (which happens all the time), yet I also don’t feel like I’m in “infinite distraction” desktop mode as when I am sitting at my computer.

messenger channel with only me inside (make a group, kick your friend)

if I’m on windows: notepad++ and it’s “untitled”, unsaved documents that still get stored to disk, enabling you to have dozens of them
linux: random kate window

Keep it as simple as possible. For me that means appending to a single notes.org file in Emacs. I access this Emacs session and file from all devices over ssh. Periodic git commits.

I use a bunch of markdown files in a folder together with a script to search/edit them that utilizes fzf and ripgrep. They can contain images. I wrote a small vimscript function to make inserting them painless. To view them, I use pandoc to convert them to html and display them in a browser. That’s also done by the script. If I’m not on my PC, I’ll either use Signal’s feature to send messages to myself, or I use an audio recorder, if it’s too much to type, and send myself a note to transcribe it when I’m on my PC. Additionally, those notes live in a Syncthing folder, so I can always access them from my phone If I want too.

My “always there” option has become setting a reminder using Siri. Something like:

“Hey Siri, remind me Monday morning that the solution to the X problem is to do…”

It’s imperfect, but it fully solves the “I don’t want to lose this thought” problem for me.

I assume other virtual assistants can do the same.

I use a spiral notebook (A5 or so) and a pen. Spiral is important because you just leave it open to the next page. I write things down as needed, either capturing ideas, thinking through something, scribbling reminders.

Then, at least once a week as part of my weekly review, I process those notes. I draw a little X on a page that I finish processing. Processing means everything on the page goes where it needs to: calendars, todo lists, reference files, project trackers…

This distinction between capturing and processing is really helpful. I learned it from David Allen’s GTD.

Vimwiki in a cloud folder for sync across devices. It works well enough and has good support for code snippets etc

For making notes I tell the information to someone else. This does make retrieval from that system difficult so for that I usually use Google.

I think its worth separating out note taking / brain dumping from polished ideas.

I see no value in searching brain dumps, to me those are a spur of the moment activity which may or may not result in something worthwhile. But the process doesn’t serve much value after its done. Kind of like a first draft of a design, throw it away & rewrite it.

Another analogy, i used to bring a zoom hand held recorder to each practice / jam with one of my more improvisational bands in hope of listening back & recovering gems to improve upon - countless hours of audio, nothing to show for it. It has proven very useful however to record a rehearsed version of a song that was just finished. After which I listen back in coming days to internalize & see if there could be something improved / removed.

So going back to software, I think I care a lot more about storing tangible ideas than just ideas. To me that means prototypes, the way i do that in usually have a “experiments” folder with all my working but unfinished ideas each is usually a small project.

I do wish for a better spot for experiments, something like computational notebook for any language that is searchable etc, id love that

4”x6” mini whiteboards. I have some taped up around the house (e.g. kitchen, front door).

I will topically add things to Todoist, and I have a cron thing that will fetch a specific todo list and send me SMS and email throughout the day, but when I have something pop into my head that I don’t want to forget, and making some formal note someplace is inconvenient at that moment, I write it on one of those whiteboards. This way I will keep running into it during the day (I work at home). This is handy for things I want to do that day,

Other hack: Write stuff on paper, take a picture with my phone, email it to myself. I lose papers. :)

For software, to leave a note for myself for projects, I have a Ruby script called what.rb that manages a YAML file keyed off directory paths. I can add little notes this way. What is this project for? What was I supposed to do next? The what script will either add a new note or show the current note for that directory.

For general note taking and random stuff when I have more time: Joplin.

Joplin uses a SQLite db. I have a Ruby script that will connect to the db and grep through the notes text, looking for the text #NEXTSTEP:. E.g.

#NEXTSTEP: Finish the README!

It collects all of these and prints them to the console along with the note name so I can see where I left off on the myriad side projects I have.

Over the years I have tried so many ways to do notes. They don’t really work well for me. Pen and paper is fine, but slow, tedious, and not easily searchable.

I’ve tried Wikis, FOAM (with vs code), just txt files, Evernote, Onenote, Joplin, etc. Ive tried Zettlekasten and many other styles. Nothing has stuck.

I am using Obsidian nowadays, and it works mostly ok for me, but I just forget to open it. I have a blog and I used to do Today I Learned type posts, and i do blog posts on occasion.

What I have found is that I write more and do better stuff when I am writing for others, and not for myself. So I am currently rebuilding my blog/site from the ground up and it will have a public knowledge base. I will write my notes and stuff in markdown, and it will just publish.

Combo that with GitJournal on my phone, and I can do updates wherever. We will see how it goes. I don’t have high hopes that anything will work for me in general.

Two folders in my home directory.

1. ~/@ - contains notes as markdown files relevant to that computer and shouldn’t be synced anywhere else. Work stuff, private stuff, scratch paper that isn’t entirely throwaway, etc.
2. ~/@@ - contains global notes as markdown files. Non-sensitive stuff. They are automatically synced across computers and my phone via Obsidian.

I edit these with Vim. That’s about it. Honestly, I don’t use any of Obsidians features except for it’s excellent syncing and it’s mobile app. I feel a little bad about that since so many people love it’s other capabilities…but I mostly just want text files.

I use Nextcloud notes, via the browser on desktop and via the Nextcloud notes app on my phone. The Nextcloud instance is on my VPS and is backed up every hour by a Raspberry Pi with an attached SSD (which is backed up to another drive locally).

If I have notes that grow, I transfer them to the document section of Nextcloud.

My shopping list is a Nextcloud note shared with my partner.

In software projects, I add Markdown files which start with my 3 initials ‘jgy’ for which I have a global Git ignore rule. I have a script which copies any file like this to the same SSD as above. These files hold my Done, Current and TODO tasks for the project, plus other notes.

Notes which I want to refer to get moved to my TIL repository, which is public.

Google keep (will migrate off, someday…), google tasks (ditto), plain text files in dropbox

Standard Notes. It is simple, open source, secure, cross platform, has tagging and a sustainable business model, also it is quite nice to look at.

Currently I use Obsidian + Sublime Text with some bindings to edit markdown files. 1000+ of them.

Published to VitePress at wiki.nikiv.dev

I describe it in more detail here.

Soon though it will fully transfer to Learn Anything tool I am building.

All this is a conglomerate of various productivity blogs that I’ve lost entirely and personal trial and error. For me, Apple Notes with a logical folder structure (loosely based on the Dewey Decimal System) has yielded the highest reward vs effort for me. It syncs across most my devices and the ones it doesn’t, I can access my account via the browser. It has okay code support, but most importantly, support images (screenshots).

You can mirror your bookmarks/local file system to the notes structure too which is nice, plus having codified notes makes recall simple in shared conversations. (e.g. call notes with a certain integration partner at a certain place of work could be coded something like 311.56. Slap that in an email and never loose it)

Anything that requires hard thinking and is short of writing actual code, I use pencil and paper in a dated composite notebooks. You can think of the notebook as append only with no organization other than the dates. Once a notebook fills up, get another one.

Between these I feel like I can write and find most things pretty quickly without bothering with the Sisyphean tasks of finding the optimal productivity workflow.

Another philosophy that is pretty central to my workflow is that most things don’t need to be stored. I lose bookmarks all the time, and if a thought doesn’t come up at an opportune time and I fail to log it, that’s okay too.

Hopefully this helps, productivity systems can quickly stop being productive, so my advise is keep it really simple :)

I’ve tried Org Mode, Workflowy, and Standard Notes. The problem I find with them is that they become too full of useless info, and I barely reference notes.

I’m skeptical of wedging everything into a “One True Method” workflow, and have settled on a mix:

• Signal’s “Note to Self” with 1w expiration as a way of dumping ideas/links/files. This syncs between my phone and desktop, and the 1w expiration stops it from becoming a place I have to backup.
• Text/markdown/openoffice files on my desktop.

It’s not perfect, there’s lot of times I wished I had richer text, or easier access to files from phones, but it’s good enough for my needs.

i use obsidian, but i’m currently treating it as write-only

i used to do this with unsaved text files, since there’s a benefit from just writing out a thought. maybe someday, the linked nature of it will become useful

I use a combination of OrgMode, TreeSheets and a publicly available self hosted FossilWiki.

OrgMode for project specific notes, task tracking.

TreeSheets as a mind mapper, outliner, something that helps me focus during meetings.

FossilWiki for a personal knowledge base. I have two. One company-specific which runs locally on my work laptop, and a general one that runs on a VPS.

Wrote about it in detail over here: https://codesurfers.net/2021/10/02/personal-information-management.html

I’m too ADD to have any organized system. What I end up using is

• Notability on iPad to brainstorm/sketch
• DevonThink to capture web pages/links … I’m not using a fraction of its capabilities, but it’s easy to add to with the share-extension and syncs between my devices.
• Google Drive to store PDFs of papers/articles
• Safari tab-groups to organize currently-open pages. These are great because they sync, they don’t vanish when I close a window, and they let me categorize.

I use a git-managed directory of Org Mode files and LaTeX files. That’s it.

I created a “To consume” (watch/read etc.) list in TickTick and quick-add things to it. Then anything permanent gets added to the appropriate place in Obsidian.

Oddly enough, the second I start taking notes to remember, I’m able to remember less without the notes. Almost like my subconscious knows it can garbage collect that working memory.

My stationary setup (home/office): Emacs with org-mode and a custom minor-mode for lightweight OrgRoam-like functionality. Managed and synchronized with my server using Fossil as I find Fossil more fitting than Git for this use case: I need little to no complex operations and the automatic push/pull on most actions comes in handy.

My mobile setup: For reading, just shell and/or Emacs over SSH from my smartphone. For writing, preferably nothing connected to this system at all. My rule of thumb is that if I am not at a proper keyboard, any note will be low quality. Instead I prefer to scribble a quick disposable note on pretty much anything (usually either a physical notebook or my Galaxy Note) and then rewrite it properly when I’m back home.

I use Obsidian daily notes on my work laptop. I have a notebook and Lamy fountain pen for doodling and sketching there.

Personally on my iPhone and Macbook I use Bear to capture anything and everything and it’s absolutely great.

Over the years I’ve come to appreciate forgetting things. I have a stack of blank A4 papers on my desk with a couple of pens which I use to write down things in a hurry, e.g. when speaking on the phone. For planning long-running work and serious life projects, I use Microsoft OneNote (which I will be migrating off to paper soon-ish) where I draw mind maps and diagrams as an aid to organize different aspects of the project in my mind. For fun projects, I write a blog-post style write up in Emacs Org Mode as I’m going through the project which helps me pick up where I left off even if I don’t come back to it for months. I also have org-mode long todo lists that are things that I’d like to do one day which I’ll revisit once every few months, this is mostly fun project ideas with a sentence or two about the idea and an inspiration link. This is synced to my server and devices with git. If I think of something on the go, e.g. hear a restaurant recommendation from someone, I type it in my phone on a notes app that’s not backed up anywhere. Then I’ll go through the notes once every few weeks and check out the interesting things and delete the ones that no longer interest me. My todo list, appointment, events are in my nextcloud calendar and synced with caldav. Everything else I let my mind forego, if it’s very important, I’m sure it’ll come back; but I’ve found most things that happen just aren’t important.

I am too ADHD to keep notes. I tend to toy with problems in my head, and if I come up with something interesting I write a post on my blog, https://dotat.at/@/ There are a few posts I refer back to, and my blog has greater longevity and accessibility than other places I might have written down my ideas.

The other thing I have done is keep a log of interesting links, https://dotat.at/:/ which is very low effort. (It has been broken for ages, so the last couple of years of links are waiting to get published.) What surprised me is how useful it has been as an aide memoire with just my version of the link title, without tags or other notes, and with the simplest possible search facility.

• On laptop Emacs/orgmode/roam in dbox
• On mobile: beorg to capture, organice to read
• Quick capture/general daily todo: write on hand, notebooks, for a long time I used a scotch tape bracelet on which I’d write the daily todo list.
• Think: blackboard with chalk

Flow: usually I dump quick notes about meetings in notebooks then digest and enrich into org roam.