interface-experiments VS luhman-obsidian-plugin

At one point I tried creating what is Notion/Obsidian style system in 2013. I called it "living documents".

here's one of the last surviving screenshots of living documents version 1.

https://github.com/samsquire/interface-experiments/blob/mast...

The system accepted RDF N3 triples and it queried Jena Fuseki database to render graphs with d3. You could introduce facts into the system with the three boxes at the top. if you changed them, they would autocomplete and change the graph view. You could insert references or links into the document by typing them.

I have the code trapped in a JSBIN SQLite file. It's somewhere in here https://github.com/samsquire/jsbin It uses KnockoutJS.

Living documents v2 has a screen cast of it here https://github.com/samsquire/live-interface/blob/master/scre...

I'm not using the Zettelkasten method at the moment - there is a Luhmann plugin. If I were to use it I would use handwritten 5x3 index cards.

Maybe the Luhmann plugin can help? https://github.com/Dyldog/luhman-obsidian-plugin

This is indeed highly opinionated, but one choice in particular may block half the potential userbase:

Zettelkasten "Unique Note Creator"

- Is this note atomic?

- What will I link it to?

Stopping to ask yourself where to file things is likely to cause half of users to not capture them at all.

- - -

INTERRUPT-based WORKFLOW

The philosophy overview examples suggest this is using ISO8601 as the unique filename, which can be great, as then the notes are connected by time which will tend to group things with things worked on at the same time.

But the default workflow suggests author is using Luhman plugin for Zettelkasten Luhman IDs (effectively a `1a2b3c` nested outline numbering scheme) which require a conscious organization at capture time:

https://github.com/Dyldog/luhman-obsidian-plugin

This delights one class of filer, while breaking train of thought during capture for another class of filer.

I have not tested this vault myself, as I have my own "highly opininated workflow". The main thing I do is try to make capture and self-organization frictionless, then rely on a minimal structure (auto topic categories + time) to find or "garden" them later.

- - -

FRICTIONLESS WORKFLOW

Very roughly, I capture any manual notes of interest using Daily Notes, while capturing any web reading of interest using Markdownload extension that slurps the original page into Obsidian as a stripped down Markdown `.md` file, where filename is ISO8601 + URL Title.

I classify those using Google topic classifier v2, and recently am working on adding summaries with GPT-3 text-davinci-003, but recursively, where for articles more than 3500 tokens I'll summarize each subsection then summarize the summaries for an overall summary.

Hi there - does anyone use this plugin - Dyldog/luhman-obsidian-plugin (github.com)