Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Niklas Luhmann's indexing method was really advanced and simple. There was article in HN, but I have lost it.


https://www.ft.com/content/0de32666-b9fd-11dd-8c07-0000779fd... from https://news.ycombinator.com/item?id=13852785, perhaps, but I can't get past the ft.com paywall.

http://takingnotenow.blogspot.com/2007/12/luhmanns-zettelkas... from https://news.ycombinator.com/item?id=13851727 is accessible and describes the gist of the system.

The summary is an arbitrarily deep tree/list of nodes. The notation for index card headings is similar to the way we typically do outlines (1, 1/1, 1/1a, corresponding to the typical 1, 1.1, 1.1.a, etc.). Searching was easy as you could find cards grouped by the first index, then the second, then the third (so 1/1a is next to 1/1b, and anything that might sit between them is a child of 1/1a). So for a physical system, despite its size, it remained manageable. References to cards located in other branches allowed for an effective-cross referencing system, and could be paired with things like index (no pun intended) or topic cards which might list a bunch of related indices.

I suspect the physical limitations may have made it more manageable than an equivalent set of mind map nodes or a wiki, and the simple indexing (and consistent, never changing, once a number was assigned it was assigned) means everything is "static" once in place, but can have an arbitrary number of children to expand on it. This means no rewriting cross-references and/or indexes.

With a wiki you might be tempted to retitle (or find the perfect title) for a page, with this it's simply done. If you think of a better descriptor you can put it on a new card and include the index as the reference.


I built a website around this a while ago just to play around with, complete with index card looks.


Egged on by Luhmann, whom I read obsessively in my youth, I started a Luhmann-like hyper-text. Been doing it for a long time now. There are several implementations based on Luhmann's Zettelkasten, e.g. [1, 2, 3], but I've not used those myself, so can't comment on their usefulness. My own system is just HTML, grep and a few handwritten scripts. It's extremely lightweight.

It's been spectacularly useful for me. In my experience such a system becomes more useful with time, and with effort put in for curating. I spend maybe 2-3 hours every day with my private 'index cards'.

I guess most prolific writers in human history have been following a "write down every idea and refine it later" type system. But it must be so much easier now, with smartphones, computers, internet etc.

[1] http://zettelkasten.danielluedecke.de/en/

[2] http://www.verzetteln.de/synapsen/

[3] http://zettelkasten.de/tools/


I'm interested in your system. Do you explain it fully anywhere? To the extent that it's something that I could try?


Thank you for your interest. I'm afraid I have nothing in public.

The technology is near trivial, mostly scripts to generate HTML5/CSS on demand, and grep for searching. The magic comes from how I use it. Short summary.

- Write down every idea, and every insight. Write down half-baked ideas. You can later refine it.

- Don't censor yourself. (That's why I don't write a public blog.)

- When writing about some subject X, think carefully about how to link X to other pages. When making a link to another entry, read/browse the entry and let what you read give you inspiration about looking at X from another angle. This can be rather inspirational.

- Curate carefully, I spend maybe 3-4 hours every day on your ZK. Have a long-term view: this is to be used your whole life. To longer you do it, the more useful it will become.

- Be clear what you want from this system.

- The whole system wasn't useful for the first couple years. Why? Because I could still remember what I had written. That changed after about a decade.


That's very interesting. Do you use it for ideas exclusively, or "personal logging"/journaling too?. I've found a program called Connected Text[0], which bills itself as a "personal wiki", and. I'm also looking at refining my use of OneNote Desktop which I've been using for years, and it has a [[link]] feature.

[0] http://www.connectedtext.com/


That sounds neat. Any links/screenshots?


Hm, I can 'defrost' it, it's not up any more but it is something that should work without too much effort.

I put some extremely un-secure features in it (ability to evaluate code on an index card) so it never was public.

Let me see what I can do.

Edit:

Ok, here you go:

https://jacquesmattheij.com/3-x-5.png

That's just two cards, one with an apple pie recipe and one with a tiny javascript snippet but you get the idea.

3-x-5 ended up being a bit of a limitation so I made the cards auto-size as you put more text on them. I thought it would be a funny interface to a hypercard like stack with javascript as the language but I didn't pursue the idea any further.

What I did end up doing was butcher it and turn it into a tool that I use daily to do my work, but it lost the index card looks.


I take it you're referring to this Financial Times article [0] that was posted to HN a couple of days ago [1].

It's really interesting to see early solutions to things we take for granted with modern technology.

[0] https://www.ft.com/content/0de32666-b9fd-11dd-8c07-0000779fd... (Access it via Google to get past the paywall.)

[1] https://news.ycombinator.com/item?id=13852785


I've posted a few articles, look through my submission history (last weeks should suffice) if it's there.




Consider applying for YC's Winter 2027 batch! Applications are open till November 2.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: