I wasted months, just to realise the tools or systems do not matter at all. Interlinking notes doesn't scientifically make you a better writer, writing does.
Niklas Luhmann (who invented the Zettelkasten) invented his system because he was an insanely productive scholar, not the other way around.
If anything, I think re-writing works better than interlinking notes, because it forces you to consider the whole. But at some point some things becomes static enough that putting it aside and linking to it may be worthwhile.
It can't fix "you're a bad writer."
And maybe isn't even useful if you have really good memory.
But no need to posit "magic" for an improvement.
Org-mode is a combination of note taking, todolist, jupyter notebook, & calendar major-mode for Emacs. It is wonderful and terrible in a lot of the ways emacs itself is. It is incredibly configurable and extendable, and feature rich to the degree that discovering spreadsheet support after using it for 6 months seems to be a rite of passage.
Org-roam is an extension ontop of org-mode that creates an roam like layer.
The largest drawback stems from it being so flexible making it hard to develop a good phone client for it. Through Álvaro Ramírez's PlainOrg(iOS) & Orgzly (Android) do help somewhat.
Mobile-wise, lately I've been using emacs through termux on my Android (LineageOS) phone. With a keyboard attached through usb-c there isn't much difference. Even using the touch keyboard it's better than the alternatives.
If I have no connectivity, the last thing I want is to be editing my org files on yet another device and introducing conflicts. I use something like beorg (ios) for background-sync offline read-only access but don't use it to write offline.
1: https://tecosaur.github.io/emacs-config/config.html 2: https://blog.tecosaur.com/tmio/ 3: https://github.com/rougier#emacs-hacking
Apps like Logseq, Obsidian and Dendron don't have subscriptions.
The forgiving format (entries are not pre-dated so you’re allowed to miss a day, the bar on entry length is set low by design) coupled with absence of friction when traveling (no poor connectivity, low battery, etc.) really went a long way.
I knew that even the need to find a pen to write the day’s journal would be enough friction for me to skip it, which is why I tied one to it.
I’ve filled up a couple of these little notebooks, but I haven’t been able to replicate this consistency with any digital solutions.
[0] #HashtagADay: A Hashtag Journal https://www.amazon.com/dp/145213927X/ref=cm_sw_r_cp_api_glt_...
Perhaps the greatest benefit I've found in using logseq is in organizing the scientific literature I'm reading. At any point while writing, linking a paper I've read (the PDF itself, as well as my notes on it) is a few keystrokes away. This linking allows me to externalize the context with which I know a paper. I just need to browse the backlinks to see all the ways in which I've made reference to it over the entire history of my notes.
Kinopio is also such a delightful tool. I wrote about it a while ago [2] in my network thinking newsletter, source/target
[0] https://github.com/cjlm/send-to-workflowy [1] https://workflowy.com/ [2] https://sourcetarget.email/editions/15/
There was always _something_ missing from those apps. If you have time and inclination, Emacs can be customised to behave almost exactly like you want it.
Data freedom is a biggie for me. I like to have a folder of text files that I can move around, backup as I want and do fun stuff with[0].
[0] e.g. I've been having fun with making a Mastodon toot-queue org file. I just dump ideas on the org file and a cron job toots one every hour until the queue is empty.
I've been interested in coming back to remnote, the mobile app seems a lot better. The key thing I wanted in it is the spaced repetition flashcards. I feel like linking that to my notes would be really useful, although I've not got back into it and stuck with it. The best thing being the one you use I guess.
I find this approach quite similar to TeuxDeux in the OP.
Especially if someone is looking for a non-subscription Bear alternative.
The topics come in fairly random order. I keep a table of contents at front, and compile an index at the back as I go. After trying this system for a few months, I can attest that an index is, indeed, a useful thing. The great things about an index is that it allows you to find information based on multiple viewpoints.
The acid test of all these systems is "how useful have they been in practise?" As a lady in QA once sagely said: add value, not process. Or, as another poster noted, we have a tendency to fetishise different systems. As programmers, we're likely to be more guilty of this than most. But remember the whole point of the exercise. We're here to find information, not obsessively adhere to a system.
So that I can search for strings (bonus points for wildcards or regexp) in all the files in several folders and then lets you open the file at that point.
It'd be great to standardize the interop with note-taking services so people like us can write applications once and easily have them work with all the note-taking platforms out there. E.g. I'd love for someone to be able to write a spaced repetition app once, and immediately have it support any of the dozens of note-taking apps without needing special handling for each one.
I also initially went to Bear before trying out Notion and am now happily using Obsidian, which just got a Live Preview mode.
Kinopio looks great I will take it for a spin soon.
The obvious solution is to keep files in dropbox or google drive. This handles sync but not backup. I'm not sure about dropbox, but I know google drive is a nightmare to backup. You can use a few paid services to backup google drive, but you then need some kind of alerting to know when it's not running. Having to rely on a cloud storage service and a cloud backup storage does not give me long term confidence. I want something that works on most devices, is as future proof as possible and keep me in control of my data as much as possible.
Another option would be to use git to store notes, but I ran into problems when I tried to set up an auto commit and auto push. I couldn't get it working reliably. Some kind of auto pull would also be needed. Git just isn't designed to work like this so it seems like any setup would be janky at best.
Some options I've considered:
- Evernote: cloud based. rigid structure. Bad performance. I'm not in control of my data.
- emacs: I don't use emacs any more, so it's not already part of my workflow. Bad mobile experience. Same local syncing issues.
- Trilium: Really complex for someone who does not want to work with javascript. It might be too flexible for me.
- Roam: no web app last time i looked. No self hosted version.
- Logseq: Same issues as roam. Not server based / self hosted. The FAQ did mention a future option to maybe self host the optional server sync config, but it looks like it's not even on the roadmap.
My ultimate solution would be this: - Web based. This will give the ultimate flexibility for me. Day to day I either live in a browser or VSCode. An optional app would be great, but not required.
- Ideally it would save files in simple, directory based, markdown files. This would be optimal for longevity. A second best would be storing it in a Postgres database.
- Easy to backup. Flat files or a database can be easily dumped and pushed off site with a cron. I use an easy dead mans switch / healthchecks.io alert if a job has not been fired off recently. The cron does some simple checks to make sure the data is a reasonable size (above 1MB). In an ideal world, if I used a DB I would dump the contents, restore it in docker somewhere and query a test entry before pushing it off site (encrypted s3, backblaze, or both).
- The right mix of simple, but feature rich. This will be different for everyone, but for me, the key things I want are: tags, search, task lists/todo lists, directory based structure and an auto daily notes template. I'd love a button that I could click to auto generate todays notes, with some pre-set todo tasks at the top, all in a specific format. Trilium could do this, but I'm not a javascript developer, so it was a non-starter for me.
- Docker based would be ideal. I run everything I personally self host in docker. It's not a deal breaker, but I would try to implement it my self in docker if I needed to.
I know my needs are different to most. My ideal setup is based on my deep understanding of tech, containerization, cloud and the cloud. The reason my ideal setup does not exist is likely because throwing a few docker containers up on a server is a non-trivial task for most people, and a desirable thing to maintain for even less people.I'm still looking for my golden tool. Right now, I've actually gone back to evernote. It's "good enough" at its core task of note taking, but for someone with an eye to build a personal knowledge base over the rest of my career (20 years left), I'm still looking for the ideal tool.
It is at the end of the day a directory of Markdown files, so it's flexible with other tools. I have a bunch of things like Markdownload configured to write into various sub-directories.
> Both the desktop and web app don't and will not require a commercial license for both personal usage and company usage, as long as if the data are stored locally and doesn't use our server.
https://opencollective.com/logseq/contribute/free-tier-30673
I see this comment or some variation of it every time a new product is shared here that has such a model.
What I rarely if ever see is a proposal for how the developer should make money instead of a subscription.
Very often, especially for hosted apps, the alternative is not better.
> I see this comment or some variation of it every time a new product is shared here that has such a model.
I observe something similar, but we seem to have drawn different conclusions.
> What I rarely if ever see is a proposal for how the developer should make money instead of a subscription.
At the risk of sounding trite, I think people who find subscriptions unpalatable would be happy with anything that isn't a subscription. These could be offered in parallel, so subscription-averse consumers have an alternative to evaluate.
A one-off payment is the obvious and most simple choice -- wrap some limits or caveats around data / transit usage if needed. Freemium option may work in some cases (I think there's sufficient successful examples to validate this as viable). (Only) enterprise users pay. Tier the offerings to protect yourself from the heavy eaters.
I don't think expressing fatigue at the relentlessness of subscription-ONLY services is invalid.
My 0.02: people feel the exact opposite. People hate subscriptions, people hate not having control over something that they feel has been "purchased".
For example, people complained very loudly about not being able to buy photoshop anymore. Any time office365 is discussed, again people complain. People seem to hate subscriptions.
What is true instead is that people are forced into subscriptions because one-time purchases are just not offered anymore.
Said no-one ever...
I sell a one-time-purchase Mac app and I've had multiple customers tell me "I wouldn't have bought if it was a subscription" - and this is for an app that saves people hours a week on video editing, what would seem like the prototypical example of a "recurring need". I think people are just tired.