In reply to On taking notes and syndicating them by zblesk
Thank you for writing that! I too find it highly interesting to see how other people arrange their workflows, choose their tools and what they do with them. Often there are things that spark an idea or suggest a useful tweak to my own workflows. So thank you for making a comparison between how you work and what I wrote about how I work.
A few reactions to some of the things you mention.
My perspective on (personal) knowledge management is centered around the notion that I should have everything under my fingertips, and should be able to fully determine my own choice of tools. Tools one can preferably tweak, reshape or replace easily. I started taking notes in the early 1990’s and local text files were the most basic choice I made (and one of the few I then could make). Later convenience lured me into other things like Evernote, and Things for tasks, but I’ve returned to that basic starting point of using text files more recently.
At the core are these notions I hold:
- Local first. I’m from an era that connectivity can’t be taken for granted, and regularly work in settings where that is still true. It is also a dependency that even when it is usually reliable, probably carries a high cost if it does fail, as that most likely is in key moments (basically a version of the demo effect).
- Agency over tools. Tools must provide actual agency. A key part of it is being able to fully control it’s deployment and use, being able to tweak it etc. Tools must be smaller than us in that sense (not in a literal sense). Convenience may make me ignore this factor up to a certain point, but in the end having control over my tools always comes back up as an issue. Not having such control ultimately always turns a tool into a single point of failure. (Gmail and Evernote are prime examples to me) That drives me to simpler tools within my own scope of control and power to manipulate, and only allowing more complexity if it increases my personal agency significantly. It also means to me that tools need to be useful on their own, and more useful when networked.
- Personal tools. Tools need to be adaptable to the person using it. That makes it easier to make those tools smarter. As personal preferences can be assumed as the defaults, and personal routines are predictable to the person itself. Predictable routines plus preferences equal functions and parameters, i.e. code.
- Personal agency is always in the context of networked agency. In most settings the unit of agency is not the individual but a small group of connected people trying to solve something that is important to the group itself. Whatever tools the group uses should be within the scope of control of that same group. As a group’s notion of local is usually a networked notion, my local stuff needs to be able to connect (yet not depend on it). Distribution is important here. Centralisation is mostly to be avoided as it carries a cost in overhead, control and resilience.
Put that all together, and indeed POSSE basically becomes the prime directive for everything.
On WordPress: I used to handcode my sites, until I started using Movable Type shortly after I initiated my blog (hosted on a webserver at home). That was written in Perl which I was comfortable with having written my then employer’s first intranet in it. A decade later I switched to WordPress when my Movable Type install suddenly stopped working completely. I see you use Ghost, which ran a kickstarter I supported shortly after I switched to WordPress (self hosted on an external hosting package). By the time Ghost saw its first release I didn’t act on my earlier idea of running that on a home server. I’m not particularly attached to WP (also used Drupal heavily for other sites), and use it pretty bare bones, but it has served me well for the last 10 years. The switch to Gutenberg and blocks though has me thinking I might maybe go for something simpler.
On Obsidian / Joplin: I also use Joplin, but haven’t tweaked it like you have, I use it out of the box. It’s where my Evernote exports live, which from there I export to md files as needed. I treat Obsidian as a viewer, and Joplin too. Because of that I dislike that Joplin stores stuff locally in an sqlite database, obscuring the contents from my filesystem that way. From a viewer it then becomes an obscurer. Currently Obsidian has my sympathy, that may change, no tool is forever. So in my choices of e.g. plugins for Obsidian I avoid things that provide functionality that comes with a type of lock-in, where if you stop using a plugin part of your information disappears or is hard to get at because it was in a database not in the notes. I dislike YAML frontmatter too. For the Dataview plugin I use inline datafields (key:: value) which makes them a regular part of the note itself. Only when for some automation I need to know where to easily find a data field, I will put it at the top (but still not as YAML frontmatter).
On public RSS subscriptions: yes, I post a list of all the feeds I subscribe to. I treat them as individual’s voices (so no feeds from news outlets etc), and group them by my perceived social distance. I treat blogging and interacting with feeds as distributed conversations.
I always like reading about how other people process information and handle their notes/knowledge bases. It’s a topic I think about often.
Ton Zijlstra’s ideas are especially interesting to me because it seems we are trying to achieve similar goals, but go about it in opposite ways.