notes system
Collecting techniques for zettelkasten notetaking with obsidian and zotero
What attracted me to the note taking app obsidian was that it was a plain text (markdown) approach that was agnostic about file storage (whatever cloud host you use works fine) and that mirrored very closely how I already make my website (a Jekyll-built collection of markdown files that I currently sync over Dropbox).
There are lots of YouTube tutorials about how to do note taking with obsidian based off the ideas in how to take smart notes. I found early that obsidian can be as simple and as complex as you make it and it’s best to evolve your own system based on how you work rather than adopt someone else’s. Like most design problems you should start at minimum viable and then iterate to adapt to your actual workflow.
~~Here’s how I’m going to work, and this will get tested as I take notes and work through my MFA:
I’m adding readings into Zotero- I use the Zotfile plugin ‘manage attachments > send to tablet’ option to send a reading to my cloud host, and then I add all my notes in the iOS version of Adobe Acrobat. I make these notes the way you would do margin notes in Cornell system I’m extracting the gist of the text into ideas in my own words with a focus on my research interests. Then in Zotero I use the ‘get from tablet’ option in Zotero > manage attachments to retreive the document with all its new notes, and then the Zotero Mdnotes plugin > export to markdown option to save a markdown file with all the notes and bibliographic info attached. This I can add to Obsidian. ~~
ZotFiles is not working after Obsidian 6, use the obsidian Zotero Integration plugin and better Bibtex! updating this soon
This collection of notes represents a thread of thought through the text, and each note can be clicked to open that actual page in the text via Zotero. This thread of notes gets broken up as a bunch of individual notes in obsidian, threaded together, with a bibliographic card for the text as an entrypoint. Likewise for lectures I do the same- using the cornell system I do my main notes as a representation of the ‘text’ of the lecture, and my margin notes that summarize main points will get turned into cards to go into Obsidian, with a card of the date of the lecture and the speaker as an entry point. Additionally, insights I have become cards, linked to other cards from texts and lectures as a thread of thought, and all of these notes have a few searchable topic tags. Threads of thoughts can connect text notes in new ways, and new index ‘Menu of Concepts’ can link to the high-level topics in a thread of topics on a subject the way the index of a lecture or text do. These threads become an interconnected web of reusable concepts that can feed together into new threads that can then be turned into publications.
My website already has some information structures- at a low level there are 3d models, image galleries, videos and documents, then those atomic bits can be attached to posts then all of these items including posts can be part of a project, which is a sort of process book on a topic, then all of these items can be used in a page, which is a presentation-level finished project. Because both jekyll and obsidian both use markdown and cloud hosting, it’s a very fluid process from private resarch to public versions with little duplication of effort- sometimes it’s just moving a file from one directory to the next with a little alteration of the ‘front-matter’ at the top of the text file.
why?
I’m doing this not because I am a super organized productivity nerd, but because my brain is a long-abused piece of shit and I require some external information scaffolding to synthesize ideas and carry out logical and creative thinking. I’d already kind of been doing it with my website and my goal is to make the website the published iceberg tip of an organized and reusable personal research database and idea factory.
Something unique I’m adding to the process- I’ve talked to Andy Reitano about making inline web-instances of retroarch game emulators with specific savefiles, so that in the body of a text I can actually add playable emulation of a game at a specific point in play. Since I am studying games this will give me the ability to refer directly to a portion of gameplay as something to be actively played, a step up in some cases from the traditional embedded youtube video of a playthrough, referring to a portion of gameplay- “passing the controller” in a conversational way.
oh shit! someone at UCSC is already building this! - tenmile
Documents:
notes
