Org-brain – mind-mapping for org-mode
github.com
github.com
An ontology has more structure than a mind map, and you can apply reasoners to them. It was useful for me, but it undermined a layer of obfuscating abstractions some of the technologists had created as a firewall between themselves and the rest of the org, so I lost traction for being politically tone deaf.
Mind maps are easy to dismiss because they are a unifying abstraction created by someone's point of view, which must compete with every other point of view in the org. Once you model it, it implies responsibility, accountability, and that can disrupt the relationships that people have invested in.
The act of documenting something well also adds newly competent stakeholders to decisions, which has the consequence that it dilutes authority in the org.
In this way, mind maps can get a reputation for being naive technology solution to a problem that isn't actually a problem, because the nebulous relationships they expose in fact already work for someone. The helpful guy with the mind map just doesn't 'get it'.
Personally, I look forward to ontologies becoming more common in orgs, as they are immensely powerful for getting everyone aligned on a body of knowledge - assuming that's what one is aiming for.
The Big Hedge Fund Bridgewater uses decision-making assistants in their common tooling. I assume that ontologies come part and parcel with that package, considering that "alignment" and cohesion is one of their key motives.
You're right in that some of the data structure behind mind maps can be represented using a tree. However, you'd lose lots of the benefits. For example, trees might not be as memorable or as unique as a mind map[1].
[1]https://imindmap.com/blog/mind-mapping-for-memory-why-is-it-...
* Read Books
-- Book Title
Review
-- Author Name
-- Author Name 2
* Artists
-- Artist Name
But in org-brain I can connect all of these in a logical way so I can jump through it quickly.
Double thumbs up to Kungsgeten!
The ultimate that I've found so far is Flying Logic (flyinglogic.com).
- directed acyclic graph structure
- boolean logic (both AND and OR) and truth-value tracking
- multiple documents
- automatic layout (like graphviz/dot)
- python import/export api
Flying Logic is nice but it's expensive and non-collaborative.
I'm not sure what exists that is a next step up from that. Maybe owl/rdf stuff?
I generally think this entire field is basically sort of addled, and we're just waiting for software to catch up to brains. Our brains aren't hierarchical, they're graph-based. But since our physical world is hierarchical, there's this impedance mismatch and most of our software (todo lists, outliners, mindmaps) makes the wrong call by forcing us into tree-based structures.
Try representing a recipe's instructions accurately in something like OmniOutliner, OmniFocus, or org-mode. Meaning, where blocking steps accurately block their dependencies. You can't do it. But you can with a DAG.
As another example, you can't even represent a simple argument in anything less than a DAG. It's like a compiler.
1. Socrates is a man (axiom)
2. All men are mortal (axiom)
3. Socrates is mortal (1, 2)
4. All mortals will die (axiom)
5. All mortals must eat (axiom)
6. Socrates will die (3, 4)
7. Socrates must eat (3, 5)
You can't represent those dependencies in an outline, nor in most mind-mapping apps (since most of them don't allow a child to have multiple parents).
Gantt charts can represent those relationships, but if you only care about the graph structure and the dependency relationships, why be saddled with a huge project management application? And, most of them can't handle OR relationships anyway.
Graphviz/dot is decades old and isn't collaborative. But at least it's DAG with good layout.
Free-form "just draw any edge to any node!" apps aren't all that great either, not without automatic layout, not without respecting any logical relationships indicated by the edges.
My desire would be the same but with writing via stylus (Apple pencil specifically) instead of typing.
I take very free form notes during meetings and brainstorming and such via pencil and go back after to make a more coherent list.
For me, free form really needs a more direct analog input. Plus I can doodle at the same time.
There was a post from 2014 that was on the homepage last week [0], and it got me _really_ interested in org-mode + ssh sessions.
I keep a daily log in Markdown, as well as a list of tasks that I'm tracking, and the ability to seamlessly describe _and_ execute code locally or remotely was incredibly enticing.
I had the same reaction to that post, and I'm really excited to get more familiar with it, as it makes a great addition to some of the work I'm doing now.
I think there's something about having to externalize everything I know that confounds me. I wish org-brain could scan my brain and spit out something I could start working on.
We're pleased to give anyone who mentions Hacker News a 20% discount for the next week. To get the discount, just call us or use the chat system on TheBrain.com and ask for the Hacker News discount. (FYI, there is also a free version that is quite capable and never expires.)
Curiously, I'd just run across Jerry Michalski's mind-mapping via Doug Rushkoff's Present Shock a couple of days ago.
I hope it’s eventually open-sourced as I hope they’ve got loyal long-term enterprise customers to keep paying for support contracts and it would greatly expand the user-base.