I recently discovered Compendium .
This is basically a mind-mapping software with very innovative ideas about wicked problems and how to adress them. This is also a quite old open source soft that looks not supported anymore. (want to install it? prepare for hard work…)
However, there is many ideas in it that TinderBox team might (should) be inspired by.
In particular, they distinguish 3 kinds of relationships:
Associative: It is basically what we do by linking two notes in TB8
Categorical: This is what we do using same metadata for various notes
And Transclusive. It means basically that same note can be read in different contexts (maps): all copies of this notes have intrinsec transclusive relationship.
Transclusive might not count for 1/3rd of thinking potential but missing this is still a strong cut I guess.
I think that TinderBox has every thing to adress this question.
I would be happy to hear what you guys think about that.
I do know a bit about transclusion, and yes, it’s important. Tinderbox export draws very heavily on transclusion, and aliases are, in part, an approach to a novel transclusive interface.
And I know Prof. Simon Buckingham-Shum, who did lots of early Compendium work! (I’d forgotten the system name, so this was a nice chance meeting.)
There exists an inherent tension between fluent transclusivity, where the one thing is incorporated in many places, and direct manipulation, where the thing is a concrete object that remains where it is until you move it. That’s one of the central problems in hypertext research.
I’m glad to hear that you know Compendium and Transclusive relationships. Glad also that Tinderbox team already investigated around this question.
It looks like the question of aliases vs replicants. I understand that the way Tinderbox is built conflicts with transclusivity and that there is no simple way to have everything at the same place. However, I wonder if there is any possibility to get round the problem without having to deal with replicants. I agree that an alias and its original are linked by a kind of transclusive relation (transclusivity do not necessary means same note). If this relation can be made symetrical (no original, no alias, both have same value) and breackable/buildable a posteriori, then I would say that a dedicated alias-centered view, displaying transclusive relationships only could be enough for powerful analysis involving transclusive relationships.
Just thinking, might be not relevant at all.
I learned about transclusion from iA Writer. Fell in love the idea. Multimarkdown Composer does a great job. So one way I’m exploring using it is a directory of boilerplate text for some of our applications and their features. I use TextExpander to add the path + file inside the transclusion syntax.
Change the original def file and all documents update automatically. Delete the set file, it breaks the documents. Delete the transcluded content and the def file remains.
I may take an example to explain why transclusion is important to me.
I’m a physicist. I read books, economy for example. These two domains are a priori separate fields. However, one may find some similarities. For example, many people claims that the concept of energy in physics is, in many aspects, like the currency in theoretical economy. I then would like to explore the connexions between these two fields. I would like to see how far can be pushed the comparison. Doing this, I may realize that it is just a simple analogy, nothing more. But I also may realize that, in fact, economics and physics are more connected than initially thought.
To do so, I map the two theories in separate sides. Relations between nodes are simple logical links. I can draw the transclusive relations from one side to another but it becomes rapidly unreadable. To overcome this issue, I need a map where I could draw/break the transclusive links. In this map, nodes “transclusively-related” should be displayed near each other.
Here is another example but at this time I had not realized I was asking for transclusive relationships in TinderBox.
transclusion