Tinderbox Meetup, Nov. 30, 2024 (Video): Note Curation Strategies for Thinking within Tinderbox

[satikusala (Michael Becker)]

Tinderbox Meetup, Nov. 30, 2024 (Video): Note Curation Strategies for Thinking within Tinderbox

Level	Intermediate
Published Date	12/1/24
Type	Meetup
Tags	.replace, Action Code, Artificial Intelligence, ChatGPT, Export, IStamps, Thinking, Tinderbox, Zettlekasten, 5CKM, 5Cs of Knowledge Management, Eastgate, Identiy Praxis, Inc., Mark Berstein, Michael Becker, Tinderbox
Video Length	01:57:35
Video URL	https://youtu.be/rAekQ8b0gos
Example File	N/A
TBX Version	10
Host	Michael Becker

In this Tinderbox meetup, we held an improvised master class on atomizing notes and using various note-curation strategies for thinking within Tinderbox. We also discussed the default Tinderbox document settings for generating prototypes and places when # or @ are in a note name and how Explode handles notes with . ! and ? in their names.

Master class on atomizing notes: Bruce Gale is working on a book. He has been collecting feedback from people and using Microsoft Gemini AI—his reviewers have suggested that he curate his writing along various categorial themes, including specific life events at different age milestones and locations. In this meetup, we demonstrated how an author could, using Explode, prototypes, attributes, and action code, curate their notes and their writing by associating categorical list/set items and automatically linking these categories to paragraphs within one’s writing—e.g., life events, locations, and age. We also reviewed using the Hyperbolic view to visualize these linked associations.

This meetup provided a small glimpse into the power of Tinderbox. It points out how Tinderbox, when perceived as a toolbox for thinking, can help someone unshackle themselves from the thinking constraints one experiences when working in word processors like Word and Google Docs or adopting pre-prescribed frameworks and workflows. However, as great as the session was, it was also limited due to our time constraints. There is so much more to consider, such as applying other relevant theoretical approaches for any project, as well as specifics that could be used to help Bruce with his project.

Remember. Like Bruce, you can bring your projects to the community and meetups, and we can work through them together.

It will be fun to see how this and related conversations evolve.

[mwra (Mark Anderson)]

It was also noble of Bruce to act as an on-screen guinea pig trying a new—to him—feature. The occasional mis-steps we see remind us all how things are easier the second time around.

Knowing the name of the button does immediately hint at where it is (which dialog, menu, etc). But, that is part of learning any app and if working on your own and stuck, the forum’s here to help.

[macgarvin (Malcolm MacGarvin)]

Thanks for doing this, @satikusala, Bruce, @mwra and all, I found it really helpful. For those who have not yet watched it, among other things Michael showed how to automate the creation of links between notes (for map and especiallly hyperbolic view) from entered attributes, in this case Locations.

When working through the example (TBX file attached below the screenshots), I used Laurence Durrell’s four novels, The Alexandria Quartet, published between 1957 and 1960, Otherwise I followed the same steps shown in the video, including the little test of the Link Location function. It took a couple of hours to understand it and do something that I can now repurpose and elaborate.

Alexandria Quartet.tbx (421.0 KB)

You won’t break anything if you delete all the links within the note ‘Themes in …’ and delete the notes within the Locations and People folders. Indeed, if you do, you have a perfectly usable template for your own work, though you’ll need to watch the video to see how easy it is to use.

It’s a deliberately playful example to choose because the four Quartet novels run largely over the same timeline, but seen from the different perspectives of the characters whose lives and interactions only partially overlap. It examines the fluidity of time and the role of memory in shaping identity and relationships. Through its nonlinear narrative and revisiting of past events, the series demonstrates how the past is constantly reinterpreted in light of new insights. Memory serves as both a refuge and a source of torment for the characters, illustrating the tension between clinging to the past and moving forward. Durrell portrays time as cyclical rather than linear, mirroring the city of Alexandria’s timeless yet ever-changing nature.

If you really were working on studying The Alexandria Quartet, this would likely be only the first step. For example you might take each of theThemes (Love and Desire, Politics and Power etc,) and put them into the Resources section. Each, with the analogous preparations shown in the video for locations and people, would then appear as an attribute when you create a new content note, to which you would add existing or new terms, and the new terms would then be automatically added to these resources ready for use in other notes and exploration in the hyperbolic view, in the same way that new localities were added automatically to the Location and People folders in this example.

In other words a perfect subject for Tinderbox and the narrative of this Meetup (and a good read to boot!)

[mwra (Mark Anderson)]

It’s probably worth re-noting the fact, re use of links in Map and Hyperbolic view, that each discrete link type can be set to not be visible in views like Map and Timeline (see the Inspector’s Links tab). Conversely, the Hyperbolic view controls will list any/all link types in scope of the view and the user can then chose which are drawn, regardless of the ‘visibility’ setting for other views.

A consideration, if taking that nuanced approach might also be to consider which links need to be text links as opposed to basic links. The links for Hyperbolic view investigation of relationships can likely be (invisible) basic links.

Thus it is perfectly possible to have links that have different purpose/visibility in different views, yet to the same end of informing the user. Perhaps something to explore in a future meet-up, though I I’d suggest if so the we have demo file pre-made that is created specifically to illustrate this point. It’s more obvious once you see/know the concept.

[macgarvin (Malcolm MacGarvin)]

@mwra

we have demo file pre-made that is created specifically to illustrate this point

That would be good - showing and hiding link visibility is interesting - and I’d be happy to help write up the post and create screenshots/be a guinea pig to test comprehension, given that I’m a pretty unsophisticated user!

For me the most valuable extension of what @satikusala demonstrated here would be an extension of (in this case) the location example to allow nesting, so that

1. within the location folder, if you drag ‘Portsmouth’ to be a child of ‘Hampshire’ which in turn is a child of ‘England’, then all the notes that already have, or to which you in future add ‘Portsmouth’, have ‘Hampshire’ and ‘England’ added as well, and
2. when a ‘Reference’ note originally imported from e.g. Bookends (or Zotero, or a note from DevonThink) is dragged into somewhere within the Location folder it keeps the original fields but (in this case) ‘Portsmouth’, ‘Hampshire’ and ‘England’ are added into appropriate attributes.

For me that would be especially useful for what @satikusala creates an attribute called ‘Terms’ (in his own work) and I might call ‘Concepts’. I’d also be curious about why he created the ‘Term’ attribute, rather than using the ‘Tag’ attribute. I’m sure he has explained it somewhere, but I’ve not caught up with it yet.

I know that @satikusala also said that he had a more generic form of the example that he showed in the meeting. That would be icing on the cake, though just having the simple example is easier to learn from in the first instance.

Perhaps having taken up a whole Meeting with this, I’d understand if it were some weeks before we came back to this topic.

[satikusala (Michael Becker)]

@macgarvin

I’d also be curious about why he created the ‘Term’ attribute, rather than using the ‘Tag’ attribute. I’m sure he has explained it somewhere, but I’ve not caught up with it yet.

There is no mad science here, just a preference for terminology. $Tags is a set, system attribute. $Term is a set, a user-generated attribute. I like my attributes not to be plural. Also, $Term (i.e., for terminology) works better for my use case, creating glossaries for my writing rather than tags or keywords which are often used for search and discovery. Ultimately, it all comes down to preference/what “feels” right and learned behavior/muscle memory.

[macgarvin (Malcolm MacGarvin)]

Many thanks for explanation - I’ll try and pop into the next Meetup

[mwra (Mark Anderson)]

A further point re $Tags. It is as noted a multi-value (Set-type) attribute and best considered as a default keyword bucket. As such a number of processes may end up adding data to $Tags. By using your own ‘tag’ attribute, howsoever named, you also gain absolute control over what adds to/removes from it. Also recall every note can use a value for any attribute, so be using a user attribute that you will only use with a deliberate subset of notes you also don’t have to worry about have to filter for only values for the intended ‘tags’ vs. ‘tags’ stored in the same attribute elsewhere in the document. IOW, to borrow @satikusala’s terminology, only notes that might need a term will hold $Term values.

Also, noting your use of related tags (city → county → country) beware the ‘black dog’ effect that bedevils hierarchical tagging. IOW, does the titular black dog apply to dogs → colour → black or in Colour → black → dogs … or both?

‘Keywords’ and ‘tags’ are the most common names for general descriptive label words but all sorts of domains have their own tags and definition of what their use means. But, in nearly all cases it boils down to a grab-bag list of words. Fine in principle, but it is easily diluted when many tags get used. As Tinderbox lets you add attributes as needed, I have come to find that storing tags on discrete topics, or with discrete implications, in their own discrete tags is more tractable in the long term. Indeed, it makes search generally much easier, than searching a tag-bucket attribute (e.g. $Tags) and perhaps getting false positives to simple searches.

[macgarvin (Malcolm MacGarvin)]

@mwra

As such a number of processes may end up adding data to $Tags. By using your own ‘tag’ attribute, howsoever named, you also gain absolute control over what adds to/removes from it.

I get it now, thanks, that’s useful.

@mwra

Also, noting your use of related tags (city → county → country) beware the ‘black dog’ effect that bedevils hierarchical tagging. IOW, does the titular black dog apply to dogs → colour → black or in Colour → black → dogs … or both?

I suspect this gets to a deep philosophical point. Your two possible classifications is a valid consideration. I suspect knowing which way you want to jump may become easier to judge when you are working on a specific project.

For my work, when in Tinderbox, I often find myself pausing on the directionality of the links, and was thinking there should be a two way link (though putting in separate links is probably better).

My bigger(?) problem though, that gets to my original point,

@macgarvin

when you create a new content note, to which you would add existing or new terms, and the new terms would then be automatically added to these resources ready for use in other notes and exploration in the hyperbolic view

… is, by the time you get to thousands of notes in a project, manually putting in all the possible links is time consuming and so rather arbitrary (how much time did I have?). So being able to put in a term ‘carbon dioxide’ to various notes alternatively embedded within different hierarchies (Trends to Bend’; [means of] Bending the Trends; Polity, Policy, Politics; Ways of Seeing, Ways of Being [i.e are we personally bothered about carbon dioxide] and then (step 1) getting these automatically inserted as terms in that note and (step 2, possibly optional) generating links might both be useful options. It may be that my relatively few ‘big structural’ links gets me to where I need to be. It’s not the only reason why I use Tinderbox for this project. But unless it’s possible I won’t know.

But someone else, or me on another project, where the end result is known and one is trying to understand why it happened (e.g. Criminal Investigation or Intelligence analysis) you might well be interested in gathering masses of data rich notes, and seeing what drops out using Tinderbox.

[satikusala (Michael Becker)]

@macgarvin

being able to put in a term ‘carbon dioxide’ to various notes alternatively embedded within different hierarchies (Trends to Bend’; [means of] Bending the Trends; Polity, Policy, Politics; Ways of Seeing, Ways of Being [i.e are we personally bothered about carbon dioxide] and then (step 1) getting these automatically inserted as terms in that note and (step 2, possibly optional) generating links might both be useful options.

I’ve handled this in several ways. As noted, I have Tinderbox automatically curated and link notes. At times, I’ll also you the NLtagging so that notes get tagged automatically by the context in text.

I have also built a system where I roll related terms into a category and categories in them. This can facilitate interesting and down-the-chain categorization. Honestly, that is how I would do region>country>state>city. Happy to explore this on a meetup.

[macgarvin (Malcolm MacGarvin)]

Aha - this is very interesting

@satikusala

Happy to explore this on a meetup.

Yes please, happy to wait if you want to balance topics over time