Devonthink's awesome mind map mode (just add MindNode and stir)

I’ve started fiddling with a variation on outlining in Devonthink.

Imagine a MindNode mind map of the Hero’s Journey, or whatever beat sheet you want to start with for describing your plot. It’s just a starting point, not manacles.

Each node’s link points to a Devonthink tag of the same name as the node. Now you’ve got a nice clean 10,000 foot overview of your story. Click on a node, MindNode opens up the tag, showing you a focussed view of just what counts for that point in the story.

For extra hipster points, set the script for each of those tags to the transclusion script discussed elsewhere in this forum. Set the view mode to unsorted and drag the members of the tags to an order that makes sense.

Now when you click on a node in the mind map, you get a live compilation of the ideas that contribute to that point in the story as well as a focussed view of the ideas that matter. If something doesn’t fit, don’t delete it, just de-tag it so you can re-use it later.

Devonthink’s documents become a repository of ideas, tagged to appear in the story beats where they belong. The mind map is an overview and a table of contents into selected (tagged) Devonthink documents… It’s probably not strictly necessary to have that view, but it’s encouraging.

If you want to change the order of chapters in the mind map, drag them around in the mind map. You may decide the order of the tags in Devonthink isn’t really important if the mind map is your table of contents.

If you decide the Hero’s Journey isn’t so hot, make a second mind map with whatever structure you want. If you use new tags for the new mind map to point to, you can go back to the Hero’s Journey if you want. No ideas lost, no structure lost.

It’s so hip I found myself considering a man bun and skinny jeans. But I don’t fit in skinny jeans and my Joe Friday buzz cut just doesn’t make a credible man bun.

But the outlining idea seems to work and it’s a natural thing to save the MindNode documents in the Devonthink database so everything is wrapped up in a nice package.

Another idea is block out a story with shapes in OmniGraffle, each shape’s programmable action set to open a Devonthink link when clicked.


Love it!
I’m doing something similar with TheBrain, where certain “thoughts” (i.e. individual elements) are linked to DevonThink tags, whereas others are actually markdown items that get indexed in DevonThink.

1 Like

TheBrain is cool. It’s what I used prior to Devonthink. TheBrain’s jump thought concept is powerful.

I liked the idea when I saw The Brain years ago, but it never gained real usefulness for me. Glad to hear you two are making it work for you!

Don’t want to hijack this thread and make it all about TheBrain, but… :slight_smile:

From my perspective, TheBrain as a standalone solution lacks many essential features that DevonThink offers. If I had to choose one of the two applications, I would certainly go with DevonThink.

Luckily, DT’s flexibility makes it possible to use TheBrain “on top” of my DevonThink workflows with minimal additional maintenance. I use it specifically for my Zettelkasten (in combination with DT), whereas DT also has broader additional use cases.

Where TheBrain excels: Displaying non-linear relationships between a large number of concepts visually (“jump thoughts”). It’s also possible to designate different “link types” and have them displayed in specific colors. This makes it possible to open an “atomic note” in my Zettelkasten and instantly identify different types of relationships it has to other concepts. In my personal experience, this helps dramatically with making new connections.

Using KB Maestro macros that I came up with in combination with some scripts (created by @pete31, although not specifically for this purpose), it’s possible to display individual notes either as part of a thought in TheBrain, or as a Markdown item in DT. Switching between both applications is more or less instant. In the context of my Zettelkasten workflows, I regularly switch to DT e.g. to identify new connections via the See Also inspector.

Happy to do a more detailed write-up sometime if this is of interest to others.


TheBrain was my first introduction to note warehousing, and it was great until there were glitches. I had reliability problems and found Devonthink. TheBrain has probably solved those problems, particularly now that it’s no longer a Java application with the potential for a locally installed Java to run TheBrain instead of its formerly bundled Java.

Jump thoughts are what got me into tagging in Devonthink. My first exposure to DT was that it was ironclad reliable, and that I was a little crippled without jump thoughts.

Then, I realized I could tag two or more documents with a common tag. The reveal tag function would list what amounted to jump thoughts.

A feature that would make that kind of browsing more powerful, at least in my use, would be to have a way to reveal a tag in a new window, or to reveal the tag in the same window without shifting to the tag group - keep the original document in view.

That way I could view a document plus its tag-related brethren without losing context.


Yes, TheBrain + DT is of interest, please tell us more. (Maybe in a separate thread to avoid further hijacking of this one.)

I like the idea of TheBrain for seeing connections, but have found it to be fairly terrible for actually collecting data. And I love DT for data collection, but sometimes find its lack of visualization tools frustrating. So…

If something that looked like the brain could read a DT database, that would be amazing. Tags would appear as jump thoughts, and all you would need in TheBrain would be the plex and links to Devonthink documents.

It wouldn’t matter if you had to add new documents in DT proper. The visual navigation would be extremely nice.

I see that tags are potentially more powerful than jump thoughts. A node in TheBrain has it’s jump thoughts, and that’s very nice, but it’s only one group of jump thoughts. If you think of a tag as a jump set, then tags give you the chance to have as many groups of jump thoughts as you want.

And, if wishes were horses…


@kewms Okay, I’ll create a new thread on this as soon as there’s some time :slightly_smiling_face:


Yes, that would be awesome!

In practical terms, at least for now, I think the workflow needs to be as follows:

  • Index entire Brains folder
  • Add thoughts in TheBrain (with attachments, e.g. Markdown item or PDF). Purposefully include a data point that can be “read” by DevonThink smart rules.
  • Set up different smart rules that replicate, tag and otherwise modify items as desired based on the added datapoint.

To open a TB thought’s attachment in DevonThink, it should upon indexing have received its own item link added to e.g. the URL field (via a smart rule, so no added maintenance). The URL field can be accessed from within TB by invoking the “Get Info” window (a hotkey can be assigned).

And for the reverse case of jumping from the item currently displayed in DT to TheBrain, the TB link could be added to the Finder comments upon creating the item in TB (Finder comments are accessible via the “Get Info” window). Alternatively, it could be added to the contents of e.g. a Markdown attachment. With the help of a keyboard maestro macro, both options only require a single additional hotkey press when creating the item in TB.

Tinderbox has a “map” view and can “watch” a DT group, including a smart group. Its interface isn’t as visually slick as TheBrain’s, but it’s pretty powerful. (Warning, steep learning curve.)

1 Like

Ok, Tinderbox is on my list to revisit. I bought a license, long ago, and I think there is always some discount to old licenses for upgrades.

The problem for me with Tinderbox is that it seemed like my energies were going to be spent curating my ideas instead of using them. The allure of Devonthink is simplicity of use, precision in execution.

Maybe using Tinderbox as an alternate display tool for DT could be a good division of duties.

Thanks for the tip!

That’s where I currently am with Tinderbox. It doesn’t scale to the massive quantities of material that DT can handle, but it can be pretty useful in working with clusters of dozens to hundreds of notes. (The Tinderbox people will say thousands, but I haven’t personally tried that.)