I don’t understand what you meant by pointers in asking questions, but rewrote the post and emphasized the main sections and questions to aid reading.
The idea consists of two things: a workflow for note-taking and an organized system for archiving them in the filesystem. A workflow for note-taking means using a text editor (Emacs or vi). Archiving means storing these notes in a way that is not dependent on any application or the specifics of a filesystem implementation on a certain computer system. More specifically, I’m scaling a simple workflow of using a text-editor and plain-text notes, with markings in them about other notes and reference material, into a collection of documents with clickable links for moving between them. Assuming I can learn to do this in a way that is suitable for archiving, I would like to scale this further, compiling complex sets of linked notes from Scrivener into this archival system. In the end, it would be possible to write a short note in vi or a complex set of linked notes in Scrivener with an archival version of both ending up in the same base system.
From a workflow of text files with markings for references to HTML documents with clickable links
An example note:
date: …
tags: …
”I sure like to fiddle with my thoughts, but since I have trouble focusing, I think I’d like to look at [note_about_the_golden_girls.txt].”
In this example I’d need to copy the text of a tag or what is between the brackets and search for it using Spotlight or command-line tools. It would be nice to have a link that opens the other note when clicked on. I associate linking with HTML documents and web browsers. But I’m just a user. I don’t know HTML.
Question: Is it unrealistic to think that I could learn, in an acceptable time-frame, to format my notes as MultiMarkdown files in order to convert them into HTML documents?
To put it another way, is the nature of the problem analoguous to ”Sure Emacs looks like a beast from hell high on rollercoaster rides and lollipops, but you can use it as a simple text editor. Just pretend it doesn’t actually also play Tetris behind your back and not invite you”? Or is it more like ”So you don’t like the interface of Spotlight’s conversion calculations? No, you should not in fact read Bjaerne Stroustrup’s books on C++, but buy yourself a pocket calculator”.
An organized system for archiving notes and references in the filesystem
As I accumulate notes the problem of organizing them into a directory structure on the filesystem arises. As linked above, AmberV has written about a system of notes organized by having files in a directory layout of years and quarters. Her system depends on a strict chronology where files are stored using a filenaming convention of ”date+tag+title”. For example:
07193687-R-Digital Photographer as Chemist, Photographer, and Darkroom Tech.md
The creation date is included in the filename, because in her view the filesystem can not be trusted to record metadata reliably over transfers to different computer systems. Filesystem metadata provided by the Finder is really meant to be used for viewing activity on a set of files. For example: ”What files have I accessed or modified in the past weeek?”. This metadata is not meant to be relied upon for achival purposes.
Question: Is it still necessary or usefull to insert a timestamp in the filename? Have computer systems become reliable in preserving metadata over transfers to different kinds of systems, as will happen over the lifespan of an individual.
Assuming the above questions are solved, I would have a directory of notes with clickable links. I also have a need to store reference material in a separate set of directories. Amber wrote about an interesting system she called her Personal Decimal System, in Filesystem-based Information Management Question - #9 by Lettermuck.
Question: If someone has implemented such a system, how did you go about it? Any such system would be highly personal, but did you find a basic, objective framework of thinking about the system and a method of expanding it that can be used by anyone as a starting point? In my case for example, I thought of a basic principle:
Use as few categories as possible and as many as you need. Subtype these with the decimal notation where necessary.
Question: If you follow Amber’s system, do you use tags in the directory names (e.g R2.1, I2.2)? In my mind, this doesn’t seem to work since a book could be referenced in many contexts. I might write a note recording a thought I had that references Steven Pressfield. I’d tag the note with (R)ecord. But the book should probably not be filed under that letter since it might also be referenced from a different context. This would pose questions such as ”Why is the source material filed under Rx.y?”. Thus the implementation would seem to lead to managing the referece section of the system (creating symlinks and such) instead of pulling the reference and getting to work.
Question: macOS warns me about renaming a folder with a name such as X.Y.Z. Adding the third component might cause the folder to be seen as a single file. Do you know of a workaround? Dashes or underscores, sure, but I wonder how Amber did it. I tried to use curly braces around the name, but this did not solve the problem.