Train of Thought

Hand draw a query-able graph combining new nodes and edges and those already available somewhere outside on the web. github expand

Imagine a horizontal transparent layer to hold the mindgraph. On top of it new nodes and edges are drawn.

Pinned against the bottom of it are some outside nodes and edges. Outside nodes that are not part of the mindgraph layer hang by their edges downward into 3D space.

How deep is linear to the number of hops from the mindgraph layer, but not only straight down as to prevent overlap in the top-down 2D view on the system; nodes seem to slightly repel each other.

~

DMX

and a quote from the *Reasoning behind the idea* section of Ruud's `README.md` at github :

> The first advantage of allowing our digital assistants to help is lowering the chance that people think they mean the same thing, while in fact they mean something slightly or even completely DIFFERENT.