Incremental note-taking
thesephist.com
thesephist.com
This is so important. I feel most note taking apps approach note taking the wrong way: organization first, ideas later. Most of our notes are random and the only real organization they can fit into is time.
I have recently launched a note taking app [0] that keeps this in focus. In my opinion, "organizing notes" is often done after "writing notes" and that is how the app is built.
1. Note taking is a fast activity - you can take notes from any where in your phone, opening the app is fast, the steps to actually writing a note are frew, the note editor is uncluttered and stays out of the way, and the app works anywhere you have a browser & an internet connection.
2. Time is a first class citizen: all notes are sorted by date created, grouped by modifiers like year, week, month & alphabetical.
3. The app doesn't force you to organize your notes i.e., there's no special distinction between organized and unorganized notes. Some apps put unorganized notes in a specific folder or tag. Not so with my app. The focus is on writing notes; organization comes later.
4. Recalling notes is currently not the most amazing thing. There is full text search, however I feel like that's not enough. Things like similarity search (based on tags) or advanced filtering are yet to be implemented.
I think all these fit well into the incremental note taking system. However, that's a coincidence really.
Steps you need to take to write down a note with a Samsung Note:
0. Have it and screen off is fine.
1. Pull out the pen.
2. Don't press any buttons, just start writing on the display, the UI will awake when you go near the display with the pen tip.
3. Put the pen away. Note will be saved and screen will shut off.
Couple of points though:
It wasn't clear to me where the notes are actually stored. If I'm relying on your tool to collect my private thoughts, I would require raw access to files. Sure, there is backup & restore features, but I'd still require full content access without going through the client. The info wasn't in any FAQ. It seems to be a SQLite db inside ~/.config/Notesnook/. Ideally the note taking system would have open and documented format so others can build tools to insert and query notes, provide CLI interface, show dated notes on calendar, and so on. I understand this is not necessary the direction you want to take this product.
When importing a .md document, the first line "# heading" was set to be paragraph type with 24pt size instead of Header 1 type, and Header 1 seems to be generally missing from the pull-down menu. Other types of headings were correctly transformed.
As I type in a new heading using the markdown syntax "## subheader" and press Enter, the text is correctly converted to Header 2 type, but Pro version popup is triggered. I would call this a bug. Either don't convert to header, or don't show the nag screen.
The splash screen that advertised the Monograph feature could not be dismissed without pressing "Start publishing" button. Maybe that button should be changed to some less misleading text (it doesn't actually lead to any publishing feature).
I like the business model and I would gladly subscribe at this price. I also love that I can already run it and try everything out without even having an account. The thing is, right now the text formatting buttons are a minefield of PRO-only features. I just dislike when applications include buttons for pro features that open nag screen. If you want to keep the client slightly crippled for free users, you could at least disable (gray out) pro features. It would be much cleaner cut if only syncing service is paid, but I don't know enough about your target users and business model.
In general I would love to be able to restrict formatting UI to just markdown features. I don't want to accidentally change font size or mess with text alignment. All included options for text formatting are just noise to me and I prefer simplicity in note taking.
Just few thoughts from trying it for 20 minutes. I'm still thinking about using it seriously.
About storage: since the app is electron based, it stores everything in IndexedDB. I didn't want to use something separate because it'll add extra stress. However, your point about mentioning where things are getting stored is important. I'll make those changes ASAP.
> Ideally the note taking system would have open and documented format.
It's really just HTML. It's not stored in local files but the content of notes is actually just plain, normal HTML.
> When importing a .md document, the first line "# heading" was set to be paragraph type.
Hmm. That's probably because I turned of H1 since there's already a title at the top. Should the title be separate?
> but Pro version popup is triggered.
This can be annoying. It has been on my backlog to fix this and make it less annoying but I haven't gotten to it so far.
> It would be much cleaner cut if only syncing service is paid.
I took the opposite route with Notesnook. App functionality is limited but you get unlimited syncing to unlimited devices.
> In general I would love to be able to restrict formatting UI to just markdown features. I don't want to accidentally change font size or mess with text alignment. All included options for text formatting are just noise to me and I prefer simplicity in note taking.
Or in other words, a way to hide the toolbar. I noted it down for later.
Thank you so much for taking the time to point all these small things out. Means a lot.
For note taking app allowing only single title makes a lot of sense.
The note title came from the file name, and my H1 got demoted to Paragraph with larger text. Since md can have any number of H1's, it doesn't seem possible to convert it without loss of information. One option would be to detect if there is only one H1 and use that instead of file name for note title. If there is more than one, then convert all H1 -> H2, all H2 -> H3 and so on.
"Google Keep is the best note-taking mobile app I've ever used. Interesting thing is, I didn't realize how good it was until I tried to switch to a better app due to Keep's lack of an obvious export feature. (You can export your Keep notes via Google Takeout.)
What makes Google Keep so special?
— Write icon opens to the body content, not the title. Several other apps open to the title and expect you fill that first. Horrible UX.
— content body begins several pixels from the top. Beginning writing right at the top makes the content feel too far. Cc: Writer Plus, which has a 4.6 rating from 46k reviews.
— focus when scrolling through app includes content body, not titles only. The boxes show parts of the content when you look through the app. Again Cc: Writer Plus and its ilk.
— 'search' searches all of content body + title + links.
— instant cloud sync.
— lightweightness.
— tagging > notebooks/folders. Easier and faster tagging a note than moving it around.
— urls are active links.
— view box is relative to content size.
What It Lacks — Markdown, or something similar
— Import/export
— revision control
— hierarchical structuring
— sharing notes with non-keep users
— linking to other notes within the app
A single feature could solve all last three features it lacks: links. You could copy the link to a note and share with non-Keep users or post it in some other note to refer to that note. (Keep already shows titles and images associated with linked web URLs)."Noticed this so wanted to pitch in. Notesnook [0] solves most of these.
There is full markdown support, import from all the big note apps, you can share notes with anyone via Monograph [1].
Internal linking between notes is planned.
Google Keep is a awesome but only for taking quick notes. You are stuck as soon as you want to write something longform. Not to mention the privacy implications.
I'd say on-topic. The core of incremental reading is a highly configurable mechanism of distributed exposure to your notes over time, which is an activity the article advocates for. In addition to letting the user walk the revision path through the note maze, the SuperMemo implementation walks the extra mile in suggesting an optimal schedule of revision given constraints, priority criteria, and past performance. The tools mentioned in the article, including (inc)remental, do not suggest any schedule of revision, nor explicitly tackle the problem of measurably preserving information in long-term memory (which, in SuperMemo, is the culmination of the incremental reading process).
The word "reading" in incremental reading can be misleading in establishing a relationship with note-taking, but "reading", here, is far more than "consumption": elaboration/editing of notes is encouraged (even necessary) for brain-friendly formulation and memory preservation, which is also another point the article touches on (though it ultimately seems to focus more on history preservation). An elaboration-first approach to incremental reading is incremental writing[2].
Because these reviews are distributed (it is a separate algorithm to the one governing core Q/A repetitions) and also interleaved with other material, you are encouraged to operate on portions of texts on each review, giving your brain a chance to consider different successive improvements, such as:
- transformation into smaller portions (which are themselves registered and subject to distributed review),
- introduction of supplementary material (e.g. other articles, multimedia),
- better wording,
- contrasting with other sources,
- personalization.
It beats the single exposure to material and upfront "ankification" (formulation into Q/A cards), in which biases—such as single-perspective bias, or recency bias—are prone to manifest more strongly in the formulation of your material. Furthermore, because the sources of information tend to remain unreviewed after this (i.e. kept off-system), there could also be a tendency to accept their premises or claims without competition from other sources.
Issues I can find with it usually arise from the desire to meet tight deadlines: it is counter-productive to distributed reviews, crystallization of knowledge, and memory consolidation, to rush the process. The SuperMemo implementation also has a vast toolset, which may be hard to grasp in the beginning. This article[1] provides a coherent explanation of the process from a user perspective.
Edit: SuperMemo 15 and above are incremental-reading-enabled; there are freeware downloads and trials should you be inclined to try it[2].
[1]: https://www.masterhowtolearn.com/2019-08-06-supermemos-incre...
Roughly the process goes: Import article -> read section (create extract/readpoint (a bookmark of where you were in the article) -> review extracts/articles -> create cloze deletion/card (if necessary).
I can't use any other app for this process because you can't really schedule material like that in Anki/Polar Bookshelf/RemNote. Supermemo is unique in that the material you want to make cards of eventually gets scheduled itself, so you don't have to remember to read something. I would use Polar Bookshelf, but every time I've checked in on it, they haven't added that feature. In Supermemo can import dozens of articles at a time, assign them priorities, and know I'll eventually be shown them to read.
I have a couple thousand articles imported at this point, and I never need to remember to read them. This is the most important feature for me, and why I've continued use Supermemo for so many years rather than switch to Anki or Polar Bookshelf.
Not needing to remember to read something you found interesting, rather than just save it as a bookmark somewhere (every bookmarking app) is a killer feature for me, but apparently is not very useful to other people. I've tried explaining it to several people (including the developer of Polar Bookshelf), and they just don't really care.
The longer I've used Supermemo, the fewer cards I end up making from material (most things aren't really worth memorizing/remembering or even reading once you step away from it for a few days/weeks). But, I still import a lot of articles/comment threads from HN etc I eventually want to read.
Often times I'll create extracts, and not actually make cards of them, because I just like rereading that particular paragraph.
I have tasklists of business/programming/etc ideas, that also get shown during my daily reviews. I can add to them, or generally get reminded that I was thinking about X at some point. Very underrated feature.
If you don't want to see something again, you can dismiss it from the review process so it doesn't show up in your reviews. However, the information is still in your collection, so you can still search for it.
Most people take notes and then never look at them again (this comes up in every note taking thread on HN). Supermemo continuously shows me my notes over days/months/years. It also shows me interesting articles I want to read on a daily basis. It also helps me remember whatever I want. All I have to do is open the app once a day and hit next repetition.
> Captured ideas are better than missed ones.
> Ideas that can’t be recalled are worse than useless – effective search and recall form the soul of great notes.
In thinktype, you write and search notes at the same time. You can start both writing and searching (which are the same) right when you open the app.
> Time is essential to how we remember, and should be a first-class concept in a good note-taking system.
In thinktype, you see a relative time for every note ("two weeks").
When you open the app you see all notes sorted by recency. You can also easily sort search results by recency, for example to see how your thinking on a topic developed.
There's never any reason to think about where to save it, how to save it, what tags I should use, what it's linked to, formatting and other distracting tasks. You just run `notes` and it opens this month's file in your default $EDITOR, `notes this is my 1 liner to append to the file` or `echo "something to pipe in" | notes` depending on what you want to add.
It's just 1 continuous thought process of notes that's broken up by date and having it broken up by month is nice because if you're working on something that spans a few days you have everything together most of the time to recall it when you want it most.
Then for recalling specific older entries it's really easy to use grep and other command line tools or just open the whole directory in your editor and casually browse your notes if you want to go down memory lane.
- The article claims that the forgetting of 90% is an underrated problem. I think it’s generally a feature that enables us to converse, read stories, and live. It’s a garbage collector. More specifically, however, it gets in the way when we learn new things. You read something important, have an aha moment, but then forget it a minute later when your short-term memory gets refilled. And so you need to run into the same important thing multiple times before it really sticks with you.
- The article claims that the time element is of upmost importance. I think it’s important only in very specific cases, and generally unimportant. The passing of time also makes the value of it decay.
- The article claims that new notes are better than re-working old ones. I feel the opposite is my preference. You re-work your knowledge, review, correct, and restructure it. What matters is your current understanding, not how you arrived there, or what your 0.1 version was. The art is in simplification and distillation over time, not accumulation.
Would prefer a workflow where a note can be corrected and refined over couple of days, and then enters read-only mode. The only way to correct and append to it would be to create a new note with better text, and optionally redirect any links to the new note.
The author dances around the two main properties that all good note-taking methods have:
1. It's easy to add new notes 2. It's easy to look through old notes to find what you want
Pencil and paper, or bound journals, do the former very easily. The latter is much harder, unless you start some indexing system (such as a bullet journal solution). This suggests searching and tagging are ideal, which digital is good for. But, and I cannot stress this enough, (1) is more important than (2). It is not impossible to find information in hand-written notes, just harder. But the exact moment it becomes hard to take new notes, the solution fails because you won't use it. And I believe Workflowy does all of this, with quick nesting, easy extension, and even quick reorganization. And you can quickly drop into "just let me write some notes," then go back and organize later (even splitting your notes up into subcomponents, if you want).
* no easy way to insert code snippets
* because the desktop app is just a web view, it’s not possible to open more than one window. Which means that for some poweruser stuff you have to open multiple browser Windows of the web version.
im using dyanlist now which is similar enough, just more options, but i can't see myself moving away from outliners ever again
Currently, however, I'm drifting away from the journal format to a Zettelkasten, which I keep in Obsidian, a plain-text / markdown desktop app. I guess I could keep my Zettelkasten in Workflowy, but I rather like the dedicated desktop app approach for it.
workflowy is great but dynalist has a leg up here with its 'move' feature. it allows you to write a note anywhere and then search for another note to move it to. its useful when I'm in a completely different context, like if I'm in my todo list section and i want to write down a idea in my programming section, i don't have to go to that location and then back
I agree w/ most of OP, a few things I think should work differently. One to call out -- images. When I build prototypes without drag and drop image support, I find I quickly stop using it until I add support for them. It's just too useful on a computer to grab a screenshot of something or make a little diagram. This is especially true when I am in "thought dump" mode.
I also played with append only and while its _usually_ sufficient, it is very annoying when not. I compromised by making it append by default, but easy enough to right click -> edit.
This is a good write-up. Not the first one mentioning this, but ideas first, organisation second, that's the natural workflow. Yet most note-taking apps have it backwards. The self-promoting ones in the comment section seem no different besides offering a new color theme.
The only app I used that was true to this was apple notes. But then, organisation sucked and my workflow is not just about taking notes. It's about organising my knowledge and time around it.
I'm going to be no different here and promote https://acreom.com which launched recently and is unique. It's simple just like apple notes, the functionality is either well designed or hidden in the backend using machine learning.
It's time for a new approach. I'm happy to share access and get ridiculed if someone doesn't find it so.
Further, some minimal command line app would truly suck for me. What if I want to draw a diagram/sketch, express something that doesn't neatly fit into some text? Maths symbols? I don't want to type stuff. Embedding links? References? Screenshots?
An ipad pro has been a boon, I use notability or apple notes and just organise by date. What this is lacking is a good query feature.
The end game would be a clean infinite paper interface but still maybe something that I can go query on my desktop later. It should be able to at least parse maths symbols and convert it to sensible text. Diagrams it could stick into a picture.
It is very powerful but at the same time slow. Every time emacs generates the latex bits it takes some time to generate and open the file.
There is also a little bit of overhead because it is a lot of elements and a lot of syntax to generate documents.
I am learning math so my org files are full of latex code, screenshots and graphs.
It is very powerful but I wonder if I am wasting my time into details.
Sometimes I think I should limit myself to text only, but I am not sure I can convey textually ideas that are better represented in picture.
Also, although I very much like emacs I find it very slow.
So...there is this tradeoff, either very complete and detailed but slow (emacs/orgmode/deft/latex) or very basic and simple but quick (neovim/plain_txt/fzf).
I still do not know what would be better.
I don't quite understand what's the problem here. You capture ideas digitally, and later, you can revisit them just the same as you revisit paper notebooks.
I'm sure problem most people face with digital note is discoverability, things that digital Zettelkasten method want to solve. Once it inside your app, it rarely comes out. You forget about it completely except from some of most-used notes or unless you have a strong system in place(which most people don't)
Also, what paper has and digital notes lack, is spatial connection between you and your note. You can guess or estimate where or at what position your note is in your notebook, complete with time context, while digital notes live in abstract oblivion...
It seems there's been a mini-renaissance surrounding some of these topics.
github + markdown + cross-page linking works really great with his system.
I have `scratch` where I'll write just about anything, it has upcoming tasks, my current place in TV shows, juggling patterns, math problems, and drafts (including this message). Many things in it are unlabeled and I often delete things from it, relying on Git to store the history. I have `viewed-topics` which lists things (HN/Reddit threads, papers, talks) I've read/watched with a summary, my thoughts and related things from the past. Last I have dated notes (files named `YYYY-MM-DD`) where I write things that happened, philosophical rambles, mental constructions, shower thoughts, etc. Often these will reference previous notes.
I've been thinking about how to improve this for a while because it doesn't handle "updates" to old notes well. I currently have three ways I do it: (1) copy then change (with a reference), (2) append in place with a marker (WRITTEN YYYY-MM-DD), or (3) new note with just changes (and a reference). Each has problems, (1) makes seeing changes difficult, (2) is impure and can make it hard to find by date, and (3) makes seeing the "current" version difficult. I relate this to version control, particularly the tradeoff between (1) and (3). Git allows seeing both diffs and states fairly easily, but it doesn't track bits of text between files well enough. I've thought about it off and on but always manage to confuse myself without producing any solutions.
TL;DR I greatly agree but want better updates
Do you have any need to include drawings to illustrate some of the concepts?
How do you manage the math symbols and graphs, if you have any?
I don't have a good solution for graphs/pictures/drawings. I don't think I've ever needed to, if I did I think I would just put the image in a folder alongside the note (`YYYY-MM-DD-ext`).
[0] https://github.com/CoderPuppy/.files/blob/master/XCompose
Which key are you using as Compose key?
There are a few very simple features missing (which could totally be implemented via plugins)
1. A daily note page. When opening Joplin I would like to start typing, and not create a new page first
2. A button for adding Notes on top of the stack, or below/above the current note.
I love that Obsidian just stores plain markdown files, which means zero vendor lock in. And just after playing around for a few minutes I see so many things they do better than Joplin.
I can actually use it in parallel with another markdown editor (VS Code with Markdown Preview Enhanced or Typora)
https://play.google.com/store/apps/details?id=net.andrewt.qu...
This single document also contains my diary / log for daily activities.
Plain text is enough for me.
I’m currently using Craft. Previously I used Simplenote and, before that, Notational Velocity (and nvALT). My standard for notes now is that it needs to be fast, it needs decent search, and its data needs to be portable.
> 1) Captured ideas are better than missed ones
> 2) Adding new ideas is better than updating old ones. Updating notes in-place is inherently lossy, and I think it’s unnecessary.
Yes. And, setting the intention to "add," not "update" reduces friction to capturing.
When you write something in a "source of truth" - like a wiki - you've now made a commitment to keep it updated. (It's a commitment you probably don't want to keep). So, what do you do? You don't write stuff there. Instead you write it somewhere more ephemeral like Slack or a scratchpad.
Setting the expectation that your notes are to be "added to," not "updated" lets you add notes with more ease.
> 3) Ideas that can’t be recalled are worse than useless > Regardless of how you recall information back from your notes, a great note-taking system should make it trivial to get ideas out, as well as in.
Two things I'd add here are
i) It should be easy to deliver ideas to the right location (not just to recall them).
This means copy-paste should be fast. Your notes should use integrations with the tools you use to make it easy to get them there in a snap. If there's a lot of friction to getting notes organized or shared, you're less likely to do it. It's also just annoying to have to highlight text, pull it apart with your mouse, open the app you need, and paste it in.
ii) Most of what you need to recall is actually very simple to find. It's what you recently wrote.
Most of what's most important from personal working notes is stuff that's immediately relevant. Stuff that was probably written in the past week or so. This explains why so many people use Apple Notes. It's messy but it's fast to get things down and you usually can find what you need using the recency filter.
> 4) Time is essential to how we remember, and should be a first-class concept in a good note-taking system.
YES YES YES. And the stuff that we need the most right now is most likely to be stuff we wrote recently.
At https://bytebase.io we're building the fastest notepad for engineers. Our focus is on the notes that are most helpful to you as you do your work today.
>And, setting the intention to "add," not "update" reduces friction to capturing.
This is a subtle point I felt, but failed to capture, that you expressed very succinctly. Having the primary action be adding, rather than "keeping up to date", reduces the cognitive load of capture. As a result, we remember more with adding-based tools than replacement-based ones.
>Most of what's most important from personal working notes is stuff that's immediately relevant. Stuff that was probably written in the past week or so.
One thing that I've been thinking about, but haven't had a chance to form into a thesis yet, is the idea of our actions / workflows making sort of tacit modifications to our workspace that make the workspace more usable. For example, in a paper-and-pen world, if I had a note on a sticky note that I access a few times every day, it would inevitably end up somewhere very close to where I work, compared to notes I rarely access, which may be stuck inside books or on a far wall somewhere. This isn't necessarily because I brought it closer because it would make me more efficient or something -- it just ended up close by as a direct consequence of how I access the note frequently.
I feel like affordances/tacit modifications of our workspaces like this are much rarer in digital tools, where the default state is not to change over time. The behavior to have most recently edited notes "nearby" for example takes explicit design in digital notes. Makes me wonder what other kinds of tacit modifications to our tools/workspaces we are leaving out in software tool design.