Analog vs Digital knowledge

Not sure about my title but I was thinking this morning about the difference between a tool like Roam and something like Mentat or Tinderbox.

I’m not quite sure I get Roam. I mean I like it but it seems like they re-invented the wiki. It has some neat features like inferred, bi-directional, links and the “day notes” page and outliner-by-default approach. But essentially it’s free-form text with as little structure as you like and semi-structured if you spend a lot of time linking things together.

It struck me that this is quite an analog, continuous, approach to capturing & representing knowledge.

Whereas a tool like Mentat or Tinderbox which uses an approach of named entities with typed attributes is a more digital, discrete, approach to capturing & representing knowledge.

I’d never really thought about this spectrum of analog–digital in knowledge tools before.

Has it occurred to anyone else? Did it lead you anywhere interesting?

As I think about the implications of this I realise two things:

  1. It would be a mistake for me to try and copy the analog approach in Mentat. I can make a few moves to make it easier to capture freeform thoughts but structurally there’s no way for Mentat to become Roam.
  2. I should instead prefer to see how the tools can integrate with each other somehow. What kinds of problems are we trying to solve in both kinds of tools and how can they interface with each other. In particular, I want first-class support for accessing APIs to services in Mentat.