I like the notion of cards, that @visakanv describes, and threading them into a bigger whole.
What would be ideal, I think, is if all information could be represented as “cards”, and all cards could be easily threaded. Every book, every blogpost, every video, even songs, etc – all could be represented as “threaded cards”. Some cards more valuable than others.
In a way, a lot of what I’ve been trying to do with my personal knowledge management, notetaking, etc is to assemble an interesting, coherent, useful thread of thread of threads, of everything I care about. A personal web of data, with interesting trails and paths I can share with others.
I have a huge, sprawling junkyard mess of Workflowy notes, Evernote cards, Google keep cards, Notes, blogposts, etc etc ad infinitum. Buried in there are entire books worth of interesting + useful information. But it suffers from bad or non-existent threading, constrained by memory.
I too have a mountain’s worth of snippets, pieces, half sentences. And I have a much lower stack of postings and extended notes. Interesting stuff doesn’t get shared, because I envision a more extensive, a more ‘complete’ write-up that then more often than not never happens. The appeal to PKM above is key here for me. The world isn’t just cards, I agree with Neil, who pointed me to the posting above, fragmentation isn’t everything. Because synthesis and curation are important. However, having that synthesis in a fully different channel than the ‘cards’ from which it is built, or rather not having the cards in the same place, so that both don’t exist in the same web of meaning seems less logical. It’s also a source of hesitance, a threshold to posting.
Synthesis and curation presume smaller pieces, like cards. Everything starts out as miscellaneous, until patterns stand out, as small pieces get loosely joined.
I don’t know why Visakanv talks of threading only in the context of Twitter. Almost like he’s reinventing tags (tags are a key organising instrument for me). To me threading sounds a bit like a trail of breadcrumbs, to show from which elements something was created. Or cooking, where the cards are the list of ingredients, resulting in a dish, and dishes resulting in a dinner or a buffet.
More ‘cards’, snippets, I find a useful take on how to post in this space (both the blog part and the wiki part), and also bring more from other channels/tools in here.
(I took the photos during Breda Photo Festival, of Antony Cairns IBM CTY1 project, which is photos printed on IBM punch cards and held together with pins.)
Lovey post – thanks Ton. Lots of nodding while reading. One thing I keep coming back to is the difference between curating and filing. Tags, I agree, are a cornerstone of filing systems for snippets, cards etc. I’m not sure so see that as the same as curating, though. No clearly formed ideas why not – more just a feeling – but I suspect if tagging was curating then those tags would be focused on questions (Eg “ThisHasSomethingToDoWithAnsweringX”) rather than statements (eg ThisHasToDoWithTheTopicY). In my experience, and in my own usage, they are filed almost as if for later use if, as and when a question arises.
This is a project I started in December 2019 for myself. The goal is to create my own “ideal” bookmarking tool. If and when it works it might be useful as a service to others too. I optimistically registered a domain name for that eventuality: linqurator.eu
I see three layers to it:
individual bookmark collections,
a social layer like Delicious used to have,
and a visualisation and analysis layer.
The latter is something that currently doesn’t exist as far as I’m aware (though examples of analysing / visualising bookmark collections do exist, just not as a regular feature of a bookmarking tool)
Key elements I am taking as starting points for the design:
Can be used offline (working with bookmarks that is, not the bookmarking itself obviously)
Fully based on PHP/Mysql as most commonly available web infrastructure
Useful stand-alone
More useful if connected to others (why?)
Using IndieWeb protocols for connection and interaction where useful
GDPR compliant, and on EU based servers (why?)
First I will attempt to build something I can and want to use myself. After that I’ll decide on adding the ‘social’ part, and beyond that the visualisation / analytical part across multiple collections.
I want to approach this in a number of small sprints, otherwise it will never happen. Two week sprints, with 13 of them to get to a working prototype for myself.
The first sprint is doing a first general plan working backwards from the imagined final service to first steps.
Currently imagined elements to sprint towards:
General plan working backwards from imagined result
Basic database design
Basic API creation
Import my Delicious archive through API
Display page
First add bookmark form
First bookmarklet
Basic timeline view of bookmarks and tags
First search form
Attempt Threading, for e.g. a Linkpost, or as input for a draft blog posting
Parse original metadata / microformats of a bookmark
Put a saved bookmark in archive.org for a permanent external reference (might not work for paywalled pages)
Save the full page of a bookmark locally (to not bump into linkrot/paywall later, and for local full text search options)
I’m intrigued by Zettelkasten, that Roel Groeneveld describes in his blog. Zettelkasten means filing cards cabinet, so in and of itself isn’t anything novel. It’s all in the described process of course, which originates with systems thinker Niklas Luhmann. I recognise the utility of having lots of small notes, and the ability to link them like beads on a necklace, which is much like the ‘threading cards‘ I mentioned here recently. A personal knowledge management process is extremely important, and needs to be supported by the right tools. Specifically for more easily getting from loose notions, to emergent patterns, to new constructs. Balancing stock and flow. Zettelkasten coming from a paper age seems rather focused on stock though, and pays less attention to flow. Crucially it encourages links between notes, a flow-like aspect, but to me often the links carry more meaning and knowledge than the notes/nodes it connects. The reason for linking, the association that makes a link apparent is an extremely valuable piece of info. Not sure how that would find its place in the Zettelkasten process, as while links exist, they’re not treated as a thing of meaning in their own right. Also some of the principles of the process described, especially atomicity, seem prone to creating lots of overhead by having to rework notes taken during a day. That type of reworking is I think best done in the style of gardening: when you are searching for something, or passing through some notes anyway, you can add, change, link, split off etc.
Tossed out filing card cabinets of the Manchester City Library (NH/USA), image license CC BY SA
In terms of tools, I am on the look out for something other than Evernote that I currently use. What I like about it is that it ‘eats anything’ and a note can be an image, text, web page, book, pdf, or a drawing, which I can add tags to, and can access through scripts from e.g. my todo tool, etc. Zettelkasten is fully text based in contrast. As a strong point that means it can be completely created from plain text files, if you have a tool that allows you to create, edit, search and put them in an overview extremely fast. But very often ideas are contained in images as well, so dealing with media is key I think. The Zettelkasten tool The Archive is worth a try, but lacks precisely this type of media support. Devonthink on the other hand is way over the top, and let’s one loose oneself in its complexity. The Archive keeps things simple, which is much better, but maybe too simple.
I’ve been using Obsidian a little over 100 days now. So, with over three months of daily use it’s good to review the experience. I will do this in some detail, and it will span several blogposts. To explain both the evolution over time, as well as how I currently work with Obsidian in practice in a more detailed way, as Frank (rightly!) requested.
My system leads the use of tools
First off, a key point to make. I am using a system for myself to plan and do my work, maintain lots of things in parallel, and keep notes. That system consists of several interlocking methods, and those methods are supported by various tools. What I describe in my review of 100 days of using Obsidian, is not about Obsidian’s functionality per se, but more about how the functionality and affordances of Obsidian fit with my system and the methods in that system. With a better fit with my system and methods, I can reduce friction in my methods, and reduce the number of tools I need to use in support of those methods. At the same time, the use of a new tool like Obsidian influences the practical application of methods, it creates a different daily practice. Those shifts are of interest as well.
What I started with
The image below shows you how my overall system of work and taking in information looks. It’s a personal knowledge management system, that both takes care of the networked nature of making sense of new information and evolving interests, as well as the more hierarchical nature of working on projects and executing tasks. Both start with my general notion of where I want to be headed (‘goals’).
I used different tools for different parts of that image:
Excel (orange) for: listing goals (3-10 yrs out), the 3 month planning cycle I keep (along the lines of ’12 week year’), the habits I want to maintain or introduce, and tracking of those habits and project progress/fulfillment.
Things (red) for: areas of my life I’m active in, projects within those areas, and tasks in those projects.
WordPress (darkblue) for: daily logs (which I started keeping end of April this year, on an internal WP instance), week logs (internal draft blogposting), and of course for public blogging itself.
Evernote (blue) for: a list of all my current interests/favourite topics, all types of note taking, related to my work/projects and my information diet.
Other tools (grey) come into play for feedreading (Readkit), blocking time (Nextcloud calendar in Thunderbird), book reading (Kindle, Nova2), keeping references (Zotero since June, Evernote before that)
While evaluating my system, I tried Obsidian
In the spring I had started evaluating my system. I found I was not keeping up several parts of it, had fallen out of practice with a number of elements, and had changed some of my practices without adapting the flow in my tools. It had therefore suffered in its usefulness. Being at home because of the pandemic allowed me to allocate some time to take a better look, and to start testing some changes. On the tool side of that evaluation, I want to get rid of Evernote (as a silo and single point of failure) since some years.
One change in my system I was experimenting with, was keeping better atomic notes about the core concepts and key elements in how I work. Late last year I thought a bit about atomic notes, i.e. cards with individual snippets, and bringing those collections of snippets and the process of curating them and threading them into e.g. a blogpost or a line of argumentation. In January I came across Zettelkasten and took a closer look, in the spring I read a book about Zettelkasten and knew I wanted to adopt parts of it into my system (linking notes first and foremost, and storing references in a better way). That’s when I started using Zotero to keep references, and stopped doing that in Evernote (Zotero can take website snapshots and store them locally, something I used Evernote for a lot. On top of it if you give Zotero a reference it will find and store a PDF of a scientific article, very useful to read more deeply).
I started to keep atomic notes, sometimes called ‘evergreen notes’ which I to myself now call Notions, capturing concepts from my work (so not work related notes, but conceptual notes) first in both WordPress and Evernote simultaneously. WordPress (a local instance on my laptop, not online) because I already used it for day logs since April, and it allows relatively easy linking, and Evernote because it is much easier to keep notes there than WP, but linking in Evernote is much harder. I also played with some note taking tools, and that’s when I came across Obsidian. It immediately felt comfortable to use it.
How after 100 days Obsidian has covered my system
After over 100 days of Obsidian my use of it has expanded to include a much larger part of my system. Along the way it made my use within that system of Things, Evernote and almost Excel obsolete. It also means I sharpened my system and practice of using it again. This is how the tool use within my system, with the use of Obsidian in green, now looks
Obsidian now contains some 1200 mark down files. 500 are Notions, atomic notes almost exclusively about my own concepts and other core concepts in my work, in my own words. Mostly taken from my own blogposts, reports, and presentations over the years. The other 700 are some 115 day log / week log / month maps, about 100 proto-notions and notes that contain conceptual info to keep from other sources, and some 500 work and project related notes from conversations and work in progress. This sounds as a very quantitative take, and it is. I have in the past months definitely focused on the volume of ‘production’, to ensure I could quickly experience whether the tool helped me as intended. I think that monitoring the pace of production, which I’ve done in the past months, will no longer be relevant by the end of this year. I used the quantity as a lead indicator basically, but have been on the lookout for the lag indicators: is building a collection of linked notes leading to new connections, to more easily creating output like blogposts and presentations, having concepts concisely worded at hand in conversations to re-use? And it did. One very important thing, central to the Zettelkasten method, I haven’t really tried yet however, which is to use the current collection as a thinking tool. Because I was more focused on creating notions first.
On Obsidian as a tool
There are four things in Obsidian that are to me key affordances:
it is a viewer/editor, a fancy viewer/editor, on top of plain markdown text files on my laptop. It builds its own local database to keep track of links between notes. Whatever happens to Obsidian, my data is always available.It being ‘just’ a viewer is important because Obsidian is not open source and won’t be. There is a potential open source alternative, Foam, but that tool is not yet developed enough.
being ‘just’ an editor means using regular text files, it feels like coming full circle, as I have for the most part been note taking in simple text files since the late ’80s. Textfiles always had my preference, as they’re fast and easy to create, but it needed a way to connect them, add tags etc., and that was always the sticking point. It means text files are available outside of Obsidian. This allows me to access and manipulate notes from outside Obsidian without issue, and I do (e.g. on mobile, but also with other software on my laptop such as Tinderbox that I used for the images in this post).
it makes linking between notes (or future links) as simple as writing their filenames, which is supported by forward search while you’re typing.
it shows graphs of your note network, which to me is useful especially for 2 steps around a note you’re working on.
I use Obsidian as simple as possible; I do not use plugins that are supposed to help you create notes (e.g. the existing Zettelkasten and Day log plugin), because they make assumptions about how to create notes (how to name them, which links to create in them). I created my own workflow for creating notes to avoid functionality lock-in in Obsidian: day logs are created manually by keyboard shortcuts using Alfred (previously TextExpander), as are the timestamps I use to create unique file names for notes.
Timeline of three months of Obsidian use
Below is a timeline of steps taken in the past months, which gives you an impression of how my use of Obsidian in support of my system has evolved.
November 2019 I discuss the concept of cards (i.e. atomic notes), curation and writing output
January 2020 I first looked at the Zettelkasten method and some tools suggested for it. I mention the value of linking notes (possible in Evernote, but high friction to do)
May 2020, read the book about Zettelkasten by Sönke Ahrens, adopted Zotero as a consequence.
7 July started with deliberately making Zettelkasten style atomic notes in WordPres en Evernote in parallel, to move away from collecting as dumping stuff in your back yard. Atomic notes only concerning my concepts in my work.
8 July started using Obsidian, after having just started creating ‘evergreen’ notes
15 July having made 35 atomic notes, I make a new association between two of them for the first time.
28 July I’m at 140 conceptual notes. I named the collection Garden of the Forking Paths. I switched my digital tickler files (a part of the GTD method) from Evernote to Obsidian. I had stopped using them, but now it felt normal again to use them. The post I wrote about this, was made from atomic notes I already had made beforehand.
5 August I find I haven’t used WordPress anymore for my day logs ever since starting with Obsidian, and that I also added week logs (an automatic collation of day logs), and monthmaps (a mindmap at the start of the month listing key upcoming things and potential barriers). My Evernote use dropped to 4 notes in 4 weeks, whereas it was 47 the 4 weeks before it. After almost a month of Obsidian, I am getting more convinced that I am on a path of ditching Evernote.
12 August I renamed my ‘evergreen’ notes, that contain my concepts mostly, to Notions, as the generic word notes doesn’t make a distinction in the character of some the things I’m putting into notes.
12 August I write a first long form blogpost made from Notions
13 August Added Nextcloud synchronisation of the note files, allowing mobile viewing and editing of notes
31 August I keep track of tasks in Obsidian and drop Things. There was a time I always did such things in straightforward text files. Being able to do so again but now with a much better way of viewing and navigating such text files and the connections between them, makes it easy to ‘revert’ to my old ways so to speak.
13 September I am at 300 Notions. These first 300 notions are mostly my notions, the things that are core to my thinking about my own work, and the things I internalised over the past 25 years or so, of doing that work. I expect that going forward other people’s ideas and notions will become more important in my collection.
13 September I describe how I make notions and notes
September / October I increasingly use my conceptual Notions as reference while in (online) conversations.
5 October I gave a client presentation (about the Dutch system of base registers) pulled together completely from existing Notions.
7 October added a ‘decision log’ to my note keeping.
16 October 100 days in Obsidian, 500 Notions and about 700 other types of notes.
16 October reinstated a thorough Weekly Review (a component of GTD) into my system.
21 October I gave a brief presentation Ethics as a Practice, the second this month pulled together from existing notes.
This all as a first post looking back on 100 days of Obsidian.
Part 2: Hierarchy and Logs
Part 3: Task management
Part 4: Writing connected Notions, Ideas, and Notes
Part 5: Flow using workspaces
Part 6: Obsidian development vs my usage
This page lists some things on Personal Knowledge Management. In 2020 I’m mostly focused on processing material collected through my information strategies.
For now therefore I’m just listing some of the things I wrote recently about note taking.
Current description of my system
Hierarchical folder structure for Areas and Projects, description of day, week and monthlogs
How I do task management
Making notes and notions
How I make ‘notions’ my permanent notes
My 300th Notion
Planting the Garden of Forking Paths, a bit more about the note taking process
Second order notes, Zettelkasten
Wiki, Blogs, Note taking
Threading Cards and Zettelkasten
Threading Cards