I've got a million of them: how about named relationships?

[TimP]

Everyone is in the knowledge management business (just like everyone is in marketing and sales) but I'm actually in the business and always looking for ways to create, store and communicate knowledge - not just data or "information". UR is currently at the top of my list of flexible tools. Like I said elsewhere, if I could publish beautiful web-based renditions of my UR databases (interactive ones), it would pass from being merely useful and addictive to being classed as a drug.

This is my excuse for trying Personal Brain. It's well worth a look for anyone with lots of information to organize. The visual interface is engaging, but unfortunately incomplete. However, one feature stands out: named relationships.

In UR, the nature of the relationship between two items, whether hierarchical parent/child or logical link, is often easy to understand using the names or types of the items. For example, a folder is obviously a classifying container and the children are merely being contained as members of a class. A parent, such as "Big", would be used to classify its logically linked children without the necessarily conveying "contains".

(The beauty of UR is that this classification problem can also be expressed with a custom attribute or a colour passed around through a template. This is indicative of a powerful tool, such as a programming language, that offers multiple approaches to solving particular problems.)

ANYWAY, MY POINT: not all relationships are patently obvious, and being able to classify relationships by naming them would push UR a crucial step towards knowledge management nirvana. How relationships would be represented visually would take some experimentation. The visual tree would need an upgrade, for sure, but I don't think going to the fancy-shmancy network visualization of PersonalBrain is necessary.